Detailed Description
In the following description, for purposes of explanation and not limitation, specific details are set forth such as the particular system architecture, techniques, etc., in order to provide a thorough understanding of the embodiments of the present application. It will be apparent, however, to one skilled in the art that the present application may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.
It should be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It should also be understood that the term "and/or" as used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations.
As used in the present description and the appended claims, the term "if" may be interpreted as "when..once" or "in response to a determination" or "in response to detection" depending on the context. Similarly, the phrase "if a determination" or "if a [ described condition or event ] is detected" may be interpreted in the context of meaning "upon determination" or "in response to determination" or "upon detection of a [ described condition or event ]" or "in response to detection of a [ described condition or event ]".
Furthermore, the terms "first," "second," "third," and the like in the description of the present specification and in the appended claims, are used for distinguishing between descriptions and not necessarily for indicating or implying a relative importance.
Reference in the specification to "one embodiment" or "some embodiments" or the like means that a particular feature, structure, or characteristic described in connection with the embodiment is included in one or more embodiments of the application. Thus, appearances of the phrases "in one embodiment," "in some embodiments," "in other embodiments," and the like in the specification are not necessarily all referring to the same embodiment, but mean "one or more but not all embodiments" unless expressly specified otherwise. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless expressly specified otherwise.
In the embodiment of the present application, the execution body of the flow includes a terminal device. The terminal equipment comprises, but is not limited to, equipment such as a server, a computer, a smart phone and a tablet personal computer and the like capable of executing the method disclosed by the application. Fig. 1 shows a flow chart of a forest fire early warning method based on the satellite internet of things, which is disclosed in an embodiment of the application, and is described in detail as follows:
S101, a three-dimensional space model is established according to pose data and forest topographic data of each monitoring device, monitoring areas of each monitoring device are simulated in the three-dimensional space model through topographic maps and satellite images, and potential monitoring blind areas and potential overlapping areas are determined in the monitoring areas.
In this embodiment, according to the position coordinates and the monitoring angles of each monitoring device, a three-dimensional space model is established in combination with actual topographic data of the forest, the monitoring area of each monitoring device is simulated through a topographic map and satellite images, and potential monitoring blind areas and overlapping areas are marked. Specifically, according to the position coordinates and the monitoring angles of each monitoring device, combining with the actual terrain data of the forest, and establishing a three-dimensional space model of the forest through a three-dimensional space modeling algorithm. Registering the topographic map and the satellite image to obtain the position and the orientation of the monitoring equipment in the three-dimensional space model. And simulating the monitoring area of each monitoring device in the three-dimensional space model according to the field angle and the monitoring distance of the monitoring device to obtain the visual range of each monitoring device. And performing superposition analysis on the visual range of each monitoring device to identify a monitoring blind area and an overlapping area. And (3) for the identified monitoring blind area, performing spatial interpolation by adopting a Thiessen polygon algorithm, and estimating the terrain features in the blind area. And simulating vegetation distribution conditions in the blind areas by a Monte Carlo method according to the estimated topographic features. Comparing the vegetation distribution obtained by simulation with the actual vegetation distribution of the monitored overlapping area, and evaluating the accuracy of the simulation result. And if the simulation result has higher matching degree with the actual situation, adding the vegetation distribution obtained by simulation in the dead zone into the three-dimensional space model, otherwise, adjusting simulation parameters, and carrying out simulation again until the accuracy requirement is met. And finally, marking a monitoring blind area and an overlapping area in the three-dimensional space model, and providing a basis for optimizing the layout of the monitoring equipment.
It will be appreciated that a Beidou positioning module needs to be installed or integrated on each monitoring device, the module can receive Beidou satellite signals and calculate the accurate position (longitude and latitude) of the device, the position data calculated by the Beidou positioning module needs to be collected and transmitted to a data processing center, and the collected position data can be cleaned, integrated and matched and correlated with other related data (such as the monitoring angle, the topographic data and the like of the monitoring device). In the three-dimensional modeling process, the integrated position data will be used to determine the exact position of the monitoring device in the three-dimensional model.
The satellite images are acquired through remote sensing satellites, particularly, the shooting frequency of the satellite images can be set according to the characteristics of the satellites, task requirements and cost effectiveness, for example, for a monitoring area needing quick response, a low-resolution multispectral satellite can be adopted, the shooting frequency is set to be 1-2 days, for an area with low requirement on monitoring instantaneity, a high-resolution optical satellite, a medium-resolution multispectral satellite and the like can be adopted, and the shooting frequency is set to be within a range from days to weeks.
In some embodiments, the step S101 specifically includes:
according to the position coordinate data of each monitoring device, acquiring the accurate position of each monitoring device in the forest to obtain three-dimensional space coordinate data of the monitoring device;
based on the triangulation principle, determining the monitoring range of each monitoring device through the monitoring angle data of each monitoring device;
acquiring forest terrain data, and constructing a forest three-dimensional terrain model according to the forest terrain data by a digital elevation model and a geographic information system technology;
performing superposition analysis on the three-dimensional space coordinate data of the monitoring equipment, the monitoring range and the forest three-dimensional terrain model, and generating a three-dimensional space model through spatial interpolation and gridding treatment;
Simulating the monitoring areas of all monitoring devices in the three-dimensional space model to obtain the visual range of each monitoring device;
Performing superposition analysis on the visual range of each monitoring device to determine a potential monitoring blind area and a potential overlapping area;
performing spatial interpolation on the potential monitoring blind area by adopting a Thiessen polygon algorithm, estimating the topographic features in the potential monitoring blind area, and simulating the vegetation distribution situation in the potential monitoring blind area by adopting a Monte Carlo method through the topographic features in the potential monitoring blind area to obtain a simulated vegetation distribution situation;
Comparing the simulated vegetation distribution situation with the actual vegetation distribution situation, if the coincidence degree of the simulated vegetation distribution situation and the actual vegetation distribution situation is larger than a preset coincidence degree threshold value, adding the simulated vegetation distribution situation into the three-dimensional space model, otherwise, re-adjusting the simulation parameters to perform re-simulation until the coincidence degree between the new simulated vegetation distribution situation and the actual vegetation distribution situation is larger than the preset coincidence degree threshold value.
In this embodiment, according to the position coordinates of each monitoring device, the accurate position of each device in the forest is obtained, and three-dimensional space coordinate data of the device is obtained. And the monitoring angle information of each monitoring device is adopted, the monitoring range of each device is determined through a triangulation principle, and the forest area covered by the device is obtained. The method comprises the steps of obtaining actual terrain data of a forest, including geographic information such as mountain, river, vegetation and the like, and constructing a three-dimensional terrain model of the forest through a Digital Elevation Model (DEM) and a Geographic Information System (GIS) technology. And performing superposition analysis on the position coordinates and the monitoring range of the monitoring equipment and the three-dimensional terrain model of the forest, and generating a three-dimensional space grid model covering the whole forest through spatial interpolation and gridding treatment. According to the three-dimensional space grid model, three-dimensional space visualization technology such as three-dimensional rendering, texture mapping and the like is adopted to construct a realistic three-dimensional forest scene, and a three-dimensional forest display effect is obtained. The position and monitoring range information of the monitoring equipment are integrated in the three-dimensional forest model, so that the comprehensive and three-dimensional monitoring of the forest is realized, the monitoring blind area is judged, and the equipment layout is optimized.
And obtaining the geographic coordinates of the existing monitoring points around the blind area and the corresponding topographic feature data according to the identified spatial position information of the monitoring blind area. And generating a Thiessen polygon network containing the blind area by taking the monitoring blind area as a center and taking the existing monitoring points around the blind area as the vertexes of the polygon through a Thiessen polygon algorithm. And estimating the terrain characteristic value inside each polygon by adopting the known terrain characteristic data of each polygon in the Thiessen polygon network and adopting a spatial interpolation method to obtain a terrain characteristic estimation result inside the blind area. And judging the complexity degree of the terrain and the monitoring difficulty in the blind area according to the estimated characteristic value of the terrain in the blind area and by combining the area, shape and other attributes of the blind area. And determining whether a special terrain or an abnormal area exists in the blind area or not by analyzing the difference between the estimated terrain features in the blind area and the terrain features of the surrounding existing monitoring points, and providing a reference for the layout of the follow-up monitoring points. And fusing the dead zone topographic feature data estimated by the Thiessen polygon algorithm with the original topographic data to obtain more complete and accurate regional topographic information, and providing data support for optimizing the monitoring network.
The method comprises the steps of obtaining terrain information of a research area according to the Internet of things of a satellite, wherein the terrain information comprises data such as elevation, gradient and slope direction, and the like, and the terrain information is used as input conditions for simulating vegetation distribution. A Monte Carlo method is adopted, a large number of sampling points are randomly selected in a blind area, and each sampling point contains terrain information such as altitude, gradient, slope direction and the like. And according to the obtained distribution probability of different vegetation types, carrying out random simulation on each sampling point, and judging the vegetation type most likely to appear at the point. And obtaining the distribution area and the proportion of various vegetation types in the dead zone by carrying out statistics and analysis on the simulation results of all sampling points. According to the vegetation distribution situation obtained by simulation, the ecological characteristics and vegetation succession rules of the research area are combined, and the vegetation distribution in the blind area is further corrected and perfected. And finally, a relatively accurate vegetation distribution map in the blind area is obtained, and an important reference basis is provided for subsequent ecological protection and management.
For example, a three-dimensional model of a forest with an area of 10 square kilometers is built by a three-dimensional modeling algorithm. According to the angle of view and the monitoring distance of the monitoring device, the visual range of the monitoring device with the number of A-001 is calculated to be 8 square kilometers. After the visible range of each monitoring device is overlapped and analyzed, a monitoring blind area with the total area of 5 square kilometers is identified, and the total area of the monitoring blind area accounts for 15% of the area of the whole forest. For the monitoring blind area with the number of B-012, a Thiessen polygon algorithm is adopted to perform spatial interpolation, and the average elevation is estimated to be 850 meters, and the gradient is estimated to be 23 degrees. And simulating vegetation distribution conditions in the blind area by a Monte Carlo method according to the estimated topographic features. Simulation results show that the coverage of the arbor layer is 65%, the coverage of the shrub layer is 20%, and the coverage of the herbal layer is 45%. Comparing the vegetation distribution obtained by simulation with the actual vegetation distribution of the adjacent monitoring area with the number of C-005, wherein the coincidence degree of the vegetation distribution and the actual vegetation distribution is 92%, and the accuracy requirement is met. Finally marking 7 monitoring blind areas and 3 monitoring overlapping areas in the three-dimensional space model.
S102, determining a plurality of first overlapping areas from the plurality of potential overlapping areas according to a space geometric algorithm based on the three-dimensional space model, and determining a plurality of target overlapping areas from the plurality of first overlapping areas according to topographic relief data and tree shielding data.
In this embodiment, based on a three-dimensional space model, a target overlapping region is identified according to the potential overlapping region of the marker by using a space geometry algorithm, and optimization analysis is performed. Specifically, according to three-dimensional space model data, a space geometric algorithm is adopted to analyze the potential overlapping areas, and a plurality of first overlapping areas are obtained. And further screening the first overlapping area by acquiring topographic relief data and tree shielding data to determine a plurality of target overlapping areas. The space geometric algorithm comprises a calculation geometric algorithm and a computer graphics algorithm, such as a convex hull algorithm, a scanning line algorithm, a region filling algorithm and the like. When the first overlapping area is determined, a boundary polygon of the potential overlapping area is calculated by utilizing a convex hull algorithm through constructing boundary representation of a three-dimensional space model, and then the overlapping condition of the polygon is judged by utilizing a scanning line algorithm, so that the first overlapping area is obtained. And then, analyzing the terrain features of the first overlapping area by adopting a digital elevation model according to the terrain fluctuation data, and eliminating the area with larger terrain fluctuation. Meanwhile, tree shielding data are utilized, tree shielding conditions of the first overlapping area are calculated through a shadow analysis algorithm, areas with serious tree shielding are removed, and finally a plurality of target overlapping areas meeting the conditions are obtained.
In some embodiments, the step S102 specifically includes:
Calculating boundary polygons of the potential overlapping areas by adopting a convex hull algorithm, and judging the overlapping condition of the boundary polygons by adopting a scanning line algorithm to obtain a first overlapping area;
Analyzing the terrain features of the plurality of first overlapping areas by adopting a digital elevation model based on the terrain fluctuation data, and removing areas with the terrain fluctuation degree larger than a preset terrain fluctuation degree threshold value in the first overlapping areas to obtain a plurality of second overlapping areas;
Based on the tree shielding data, analyzing the tree shielding condition of the second overlapping area by adopting a shadow analysis algorithm, and providing an area with the tree shielding degree larger than a preset tree shielding degree threshold value in the second overlapping area to obtain a plurality of target overlapping areas.
In this embodiment, a convex hull algorithm is used to calculate boundary polygons of the potential overlapping regions from the boundary representation of the three-dimensional spatial model, resulting in geometric boundaries of the potential overlapping regions. Traversing boundary polygons of the potential overlapping areas through a scanning line algorithm, acquiring pixel point coordinate information inside the polygons, and judging whether the pixel points belong to the overlapping areas. And determining the specific position and range of the first overlapping region in the three-dimensional space model according to the overlapping pixel point coordinates obtained by the scanning line algorithm. And obtaining accurate geometric representation of the first overlapping region by adopting boundary polygon and overlapping pixel point coordinate information, and providing a data base for subsequent further analysis. And judging the size, shape and spatial distribution characteristics of the overlapped region through the geometric information of the first overlapped region, and providing a basis for optimizing the three-dimensional space model. And according to the analysis result of the first overlapping region, obtaining redundancy and inconsistency existing in the three-dimensional space model, and providing guidance for subsequent model optimization.
And modeling the first overlapping region by adopting a digital elevation model according to the topographic relief data to obtain a three-dimensional topographic model of the first overlapping region. And gradient values of all positions in the first overlapping area are obtained by carrying out gradient analysis on the three-dimensional terrain model of the first overlapping area. And judging whether each position in the first overlapping area belongs to the area with larger topographic relief according to the set gradient threshold value, and obtaining the distribution condition of the area with larger topographic relief. And removing the region with the large topographic relief from the first overlapping region by adopting a mask processing method to obtain the first overlapping region after removing the region with the large topographic relief. And visually displaying the first overlapped area after removing the area with larger topographic relief, visually presenting the topographic features of the area, and providing a reference for subsequent analysis.
And calculating the tree shielding condition of the first overlapping area by adopting a shadow analysis algorithm according to the acquired tree shielding data to obtain the tree shielding degree value of each area. And comparing the tree shielding degree values, determining the region with serious tree shielding, and removing the region from the first overlapping region to obtain the rest candidate target overlapping region. And further screening the candidate region by adopting a region screening algorithm according to the area size and shape characteristics of the candidate target overlapping region to obtain the target overlapping region meeting the area and shape requirements. And analyzing the position coordinates of the target overlapping areas to obtain the space distribution condition of each area, and determining the relative position relationship among the areas. And clustering the target overlapping areas by adopting a spatial clustering algorithm according to the spatial distribution condition of the target overlapping areas to obtain a plurality of clustering clusters, wherein each clustering cluster comprises a plurality of target overlapping areas with similar spatial positions. And carrying out attribute analysis on the target overlapping area in each cluster to acquire attribute information such as vegetation types, topographic features and the like of each area, and determining the ecological environment features of each area. And evaluating each cluster by adopting an ecological suitability evaluation algorithm according to the ecological environment characteristics of each cluster to obtain the ecological suitability score of each cluster. And sequencing and screening the ecological suitability scores of the clusters to obtain a plurality of clusters with highest ecological suitability, and determining a final target overlapping region set. And extracting the boundary of each region by adopting a boundary extraction algorithm according to the final target overlapping region set to obtain vector boundary data of each region for subsequent analysis and application.
For example, when determining the first overlapping region, a three-dimensional space model including 500 triangular patches is constructed, 20 boundary polygons of the potential overlapping region are calculated by a convex hull algorithm, and then 8 of the 20 polygons are judged to have overlapping by a scan line algorithm, so that 8 first overlapping regions are obtained. Then, according to the topographic relief data, the 8 first overlapping areas are subjected to topographic feature analysis by adopting a digital elevation model, for example, an area with the average gradient larger than 30 degrees is regarded as large topographic relief, 3 areas with large topographic relief are removed through analysis, and 5 areas are left. Meanwhile, tree shielding data are utilized, tree shielding conditions of the first overlapped area are calculated through a shadow analysis algorithm, if areas with the crown ratio exceeding 50% are regarded as serious shielding, the situation that the trees in 2 areas are serious shielding is calculated, and the situation is removed. Finally, through the above analysis evaluation and screening, 3 target overlapping regions satisfying each condition were obtained from the first 500 triangular patches through a series of calculations.
S103, acquiring monitoring data acquired by each monitoring device in the target overlapping area, comparing and analyzing the monitoring data of each monitoring device according to a correlation coefficient algorithm and a local anomaly factor algorithm, determining consistency and difference of the monitoring data of each monitoring device, and determining data interference factors according to the consistency and the difference.
In this embodiment, temperature, humidity and wind direction data collected by each monitoring device in the overlapping area are obtained, comparison analysis is performed, consistency and difference of the data are judged, and sources and interference factors of the data are analyzed. Specifically, temperature, humidity and wind direction data acquired by each monitoring device in the overlapping area are acquired, and the pearson correlation coefficient is adopted to calculate the correlation among the data of each monitoring device, so that a correlation coefficient matrix is obtained. According to the correlation coefficient matrix, the monitoring devices are divided into different groups through cluster analysis, the data similarity of the monitoring devices in each group is high, and the similarity between groups is low. And detecting abnormal values of each group of monitoring equipment data, calculating the abnormal degree of each data point by adopting a local abnormal factor algorithm, and marking the data points with the abnormal degree exceeding a threshold value as abnormal values. And judging the consistency of the group of data through abnormal value analysis, wherein if the abnormal values are more, the consistency is poor, and otherwise, the consistency is good. And performing variance analysis on the monitoring equipment data of different groups, calculating intra-group variance and inter-group variance, and judging the difference of the data of different groups according to the variance analysis result. If the inter-group variance is far greater than the intra-group variance, the difference between the different groups of data is obvious, and if the inter-group variance is equal to the intra-group variance, the difference between the different groups of data is not great. External environmental factors, such as terrain, building shielding, heat sources, etc., that cause data differences are analyzed based on the geographic location distribution of each set of data. And estimating the environmental parameter values of the unknown positions by adopting a spatial interpolation method such as Kriging interpolation to generate a continuous environmental parameter distribution map. Comparing the distribution map with the distribution of each group of data, and analyzing the influence of environmental factors on the data. And carrying out frequency domain analysis on the data of each monitoring device, and converting the time domain data into frequency domain data by adopting a fast Fourier transform algorithm to obtain amplitude and phase information of different frequency components. And judging periodic interference components in the data, such as vibration, electromagnetic interference and the like of the equipment according to the frequency domain analysis result. For data with obvious periodic interference, filtering processing is needed to eliminate the interference. And (3) synthesizing the analysis results, and judging the source of the data of each monitoring device. If a certain group of data has good consistency, less abnormal value, little difference with other groups, little influence by environmental factors and little periodic interference, the data source of the group can be judged to be reliable, otherwise, the data source needs to be further verified to check the interference factors.
In some embodiments, in step S103, the comparing and analyzing the monitored data of each monitoring device according to the correlation coefficient algorithm and the local anomaly factor algorithm, to determine the consistency and the variability of the monitored data of each monitoring device specifically includes:
calculating the correlation between monitoring data of each monitoring device by adopting the Pearson correlation coefficient to obtain a correlation coefficient matrix;
dividing a plurality of monitoring devices into different monitoring groups through cluster analysis based on the correlation coefficient matrix;
calculating the anomaly degree of each data point by adopting a local anomaly factor algorithm based on the monitoring data of each monitoring group, marking the data point with the anomaly degree exceeding a preset anomaly degree threshold as an anomaly value, and judging the consistency degree of the monitoring data of each monitoring group by analyzing the anomaly value;
and performing variance analysis on the monitoring data of different monitoring groups, and calculating the intra-group variance and the inter-group variance of each monitoring group to obtain a variance analysis result, and determining the degree of variability among the monitoring data of different monitoring groups according to the variance analysis result.
In this embodiment, temperature, humidity and wind direction data acquired by each monitoring device are acquired according to the position information of each monitoring device in the overlapping area, so as to obtain a data set of each monitoring device. And extracting temperature data in each monitoring device data set to obtain a temperature time sequence of each monitoring device, and determining a temperature data vector of each monitoring device. And respectively extracting humidity data and wind direction data in the data sets of the monitoring devices by adopting the same method to obtain humidity data vectors and wind direction data vectors of the monitoring devices. According to the calculation formula of the pearson correlation coefficient, the temperature data vector of each monitoring device is adopted, the correlation coefficient of the temperature data among the devices is calculated, and the correlation coefficient matrix of the temperature data is obtained. And calculating the correlation coefficients of the humidity data and the wind direction data between the devices by adopting the humidity data vector and the wind direction data vector respectively through the same calculation process, so as to obtain a humidity data correlation coefficient matrix and a wind direction data correlation coefficient matrix. And integrating the correlation coefficient matrixes of temperature, humidity and wind direction to obtain a correlation matrix of comprehensive environmental data among all monitoring devices, and judging the correlation degree of the data acquired by different monitoring devices.
And acquiring data of each group of monitoring equipment, and converting the data into an input format suitable for a local anomaly factor algorithm through data preprocessing to obtain standardized monitoring equipment data. According to the standardized monitoring equipment data, calculating a local anomaly factor value of each data point by adopting a local anomaly factor algorithm, obtaining a k neighbor of each data point by setting a proper k neighbor parameter, and calculating the local density of the k neighbor to obtain the anomaly degree of each data point. By setting an anomaly threshold, data points with anomalies exceeding the threshold are marked as potential anomaly values, and a set of potential anomaly values is obtained. And carrying out cluster analysis on the potential abnormal values according to the time stamp, the equipment number and other attributes of the potential abnormal values, classifying the abnormal values which are similar in time and space into the same class through a density clustering algorithm, and obtaining a set of abnormal events. And evaluating the severity of each abnormal event by adopting a statistical analysis method, obtaining the severity score of the abnormal event by calculating indexes such as the duration of the abnormal event, the number of abnormal values, the deviation degree and the like, and determining the abnormal event needing to be focused according to the severity from high to low. According to the equipment number and the time stamp of the abnormal event, the original data of the corresponding monitoring equipment is obtained, a time sequence diagram, a thermodynamic diagram and the like of the abnormal event are generated through a visualization technology, the occurrence process and the influence range of the abnormal event are intuitively displayed, and support is provided for subsequent analysis of abnormal reasons. Comparing the detected abnormal event with the historical abnormal event, judging whether the current abnormal event is of a known type or not through similarity calculation, if the current abnormal event is of the known type, obtaining a processing scheme of the abnormal event of the type, and if the current abnormal event is of the unknown type, triggering an alarm mechanism of the abnormal event and notifying related personnel to process.
S104, carrying out fusion processing on the monitoring data of different monitoring devices in each target overlapping area according to a data fusion algorithm to generate a data fusion model, wherein the data fusion model is used for eliminating overlapping influence caused between the monitoring data of the different monitoring devices in the target overlapping area.
In this embodiment, a data fusion algorithm is adopted to comprehensively process the monitoring data of different devices in the overlapping area, eliminate the overlapping influence, generate a data fusion model, and integrate the data fusion model into the front AI and the edge AI.
In a possible implementation manner, the data fusion algorithm is used for carrying out fusion processing on the monitoring data of different monitoring devices in each target overlapping area, and specifically includes the steps of judging the change trend of temperature, humidity and wind direction according to the terrain and vegetation details of the area, and fusing the overlapping data of the overlapping areas. Specifically, according to the details of the terrain and vegetation of the area, a Digital Elevation Model (DEM) is adopted to acquire the terrain factor data such as the altitude, the gradient, the slope direction and the like of the area. And obtaining vegetation factor data such as vegetation coverage, vegetation type and the like of the area through a Vegetation Index Model (VIM). And determining the temperature distribution trend of the area by using a Temperature Interpolation Model (TIM) according to the acquired topographic factor data and the vegetation factor data. And judging the humidity distribution trend of the area by adopting a Humidity Interpolation Model (HIM) according to the topographic factor data and the vegetation factor data. And obtaining the wind direction distribution trend of the area by using a wind direction interpolation model (WIM) through the obtained topographic factor data. If the overlapping area exists, the topographic factor data, the vegetation factor data, the temperature distribution data, the humidity distribution data and the wind direction distribution data of the overlapping area are fused, and a Weighted Average Fusion Algorithm (WAFA) is adopted to fuse the multisource data of the overlapping area, so that the topographic factor data, the vegetation factor data, the temperature distribution data, the humidity distribution data and the wind direction distribution data after fusion are obtained. And carrying out overall trend analysis on the temperature, the humidity and the wind direction of the whole area by adopting a Spatial Interpolation Algorithm (SIA) according to the fused topographic factor data, the vegetation factor data, the temperature distribution data, the humidity distribution data and the wind direction distribution data, so as to obtain the temperature change trend, the humidity change trend and the wind direction change trend of the area. And predicting the future change trend of the temperature, the humidity and the wind direction of the region by using a Spatial Decision Model (SDM) according to the trend analysis result to obtain a temperature change prediction result, a humidity change prediction result and a wind direction change prediction result in a period of time in the future of the region.
In some embodiments, in step S104, the fusing processing is performed on the monitoring data of different monitoring devices in each target overlapping area according to a data fusion algorithm, so as to generate a data fusion model, which specifically includes:
Carrying out filtering processing on the monitoring data of different monitoring devices in the overlapping areas of a plurality of templates by adopting a Kalman filtering algorithm to obtain filtering monitoring data;
Extracting features of the filtered monitoring data by adopting a related vector machine algorithm to obtain feature vectors of the monitoring data;
carrying out probabilistic reasoning on the feature vectors by adopting She Beisi network algorithm, judging the correlation between different feature vectors, and obtaining the correlation probability between different feature vectors;
when the association probability among different feature vectors is larger than a preset association probability threshold, carrying out fusion processing on the monitoring data corresponding to each feature vector to obtain fusion monitoring data;
And modeling the fusion monitoring data by adopting a hidden Markov model algorithm to generate a data fusion model.
In this embodiment, according to the monitoring data of different monitoring devices in the overlapping areas of the templates, a kalman filtering algorithm is adopted to perform filtering processing on the monitoring data, so as to obtain filtered monitoring data. And acquiring the position coordinates and the monitoring values of each monitoring device in the overlapping area by the filtered monitoring data, and calculating the accurate position coordinates of the target object in the monitoring area by adopting a triangulation principle. And acquiring the specific monitoring equipment in which the target object is positioned according to the accurate position coordinates of the target object, and adopting the monitoring value of the monitoring equipment as final monitoring data of the target object. And comparing the monitoring data of the plurality of monitoring devices, judging the motion trail and the state change of the target object, and determining whether an abnormal condition exists. And according to the judgment result, obtaining real-time monitoring information of the target object, and transmitting the monitoring information to a monitoring center so as to take corresponding measures in time.
And according to the filtered monitoring data, adopting a related vector machine algorithm to perform feature extraction on the monitoring data, and obtaining feature vectors reflecting the characteristics of the monitoring data. And training the feature vector by adopting a support vector machine algorithm through the feature vector to obtain a support vector machine classification model. And (3) acquiring new monitoring data to be predicted, and extracting the characteristic vector according to the method of the step (1). And inputting the feature vector of the new monitoring data into a support vector machine classification model, and obtaining a prediction classification result of the monitoring data through model prediction. And judging the category of the new monitoring data according to the prediction classification result, and determining the corresponding equipment operation state, such as normal, abnormal or fault and the like. And corresponding control measures are adopted according to the judged running state of the equipment, such as normal running maintenance, alarm prompt, emergency shutdown and the like, so that the safe and stable running of the equipment is ensured.
And carrying out probabilistic reasoning calculation on the feature vectors by adopting She Beisi network algorithm according to different feature vectors to obtain the association probability among different feature vectors. And judging the correlation among different feature vectors through the calculated correlation probability, and obtaining a correlation measurement result of the feature vectors. And determining a highly-relevant feature vector combination by adopting a threshold filtering method according to the correlation measurement result of the feature vector. And obtaining a fused high-dimensional feature vector by adopting a feature vector combination and a feature fusion algorithm. And inputting the fused high-dimensional feature vector into a classification model, and obtaining a final service classification judgment result through model training and prediction.
S105, analyzing the monitoring image in the potential monitoring blind area by adopting an image recognition technology, extracting topographic feature data, vegetation density data and tree height data in the monitoring image, and determining a blind area cause according to the topographic feature data, the vegetation density data and the tree height data.
In this embodiment, an image recognition technology is used to analyze a monitoring image in a blind area, extract data of topographic features, vegetation density and tree height in the image, determine a specific cause of a potential blind area, and determine data points affected by the blind area.
In one possible implementation, the method includes identifying the topographical features of hills and ravines in the image, determining whether the topographical relief results in a monitoring blind area, analyzing the tree distribution and vegetation type, and determining whether the vegetation is dense resulting in a blind area. And evaluating the shielding influence of the tree height on the visual angle, and determining whether the tree height causes a dead zone or not. Specifically, according to the input image data, an edge detection algorithm, such as a Canny algorithm, is adopted to perform edge detection on the image, so as to obtain edge information in the image. And (3) carrying out region segmentation on the edges by adopting a region growing algorithm, such as a seed filling algorithm, through the edge information to obtain different regions in the image. And extracting the skeleton information of the region by adopting a morphological analysis method, such as a skeleton extraction algorithm, according to the region segmentation result to obtain the shape characteristics of the region. And judging whether the region is a topographic feature such as a hilly or a ravine by adopting a geometric shape analysis method, such as a convex hull analysis algorithm, according to the shape features of the region. If the area is hilly or gully, analyzing the influence of the topographic relief on the monitoring visual angle by adopting a sight line analysis algorithm, such as a ray tracing algorithm, according to the topographic relief degree, and judging whether the topographic relief leads to a monitoring blind area. And analyzing texture characteristics in the region by adopting a texture analysis method, such as a gray level co-occurrence matrix algorithm, according to the region segmentation result to obtain texture information of the region. And classifying vegetation types of the area by adopting a texture classification algorithm, such as a support vector machine algorithm, according to the texture information, and judging the vegetation types in the area. And according to vegetation type classification results, analyzing the vegetation density degree in the area by adopting a density estimation algorithm, such as a nuclear density estimation algorithm, and judging whether the vegetation density degree causes a monitoring blind area or not. And estimating tree height information in the region by adopting a depth estimation algorithm, such as a binocular stereoscopic vision algorithm, according to the region segmentation result. And according to the tree height information, a three-dimensional space analysis algorithm, such as an octree algorithm, is adopted to analyze the shielding influence of the tree height on the monitoring visual angle and evaluate whether the tree height causes a monitoring blind area. And comprehensively analyzing the factors such as the terrain characteristics, vegetation types, vegetation density, tree heights and the like, and adopting a blind area recognition algorithm such as a multi-factor weighting algorithm to comprehensively judge the positions and the ranges of the monitoring blind areas in the image so as to obtain a final monitoring blind area recognition result.
In some embodiments, the step S105 specifically includes:
Dividing the monitoring image in the potential monitoring blind area by adopting an image dividing algorithm to obtain a plurality of image blocks;
Classifying the image blocks according to the color features, the texture features and the shape features of each image block based on a decision tree classification algorithm, and determining the environment category of each image block, wherein the environment category comprises topography, vegetation and trees;
combining the image blocks of the same category by adopting an area growing algorithm based on the environment category of each image block to obtain a terrain area, a vegetation area and a tree area;
extracting gray values and gradients of each image block in the terrain area, estimating elevation values of the terrain area by combining a support vector machine regression algorithm to obtain terrain height information, and calculating to obtain terrain characteristic parameters according to the terrain height information;
Extracting a color histogram and a symbiotic matrix of each image block in the vegetation region, classifying the density of vegetation by combining a random forest classification algorithm to obtain vegetation density information, and calculating according to the vegetation density information to obtain vegetation density information;
extracting edge characteristics of each image block in the tree area, detecting trunk positions by combining with a Hough transformation algorithm, obtaining tree distribution position information, and calculating according to the tree distribution position information to obtain tree height information;
And establishing a dead zone judging model by a logistic regression algorithm based on the terrain characteristic parameters, the vegetation density information and the tree height information, wherein the dead zone judging model is used for judging the dead zone cause of the potential monitoring dead zone.
In this embodiment, an image is segmented by an image segmentation algorithm based on the monitored image data to obtain image blocks of different regions. The characteristics of the color, texture, shape and the like of the image blocks are extracted, and the image blocks are classified by combining a decision tree classification algorithm, so that different categories of topography, vegetation, trees and the like are judged. And combining the image blocks of the same category by adopting an area growing algorithm according to the classification result to obtain a complete terrain area, a vegetation area and a tree area. For a terrain area, the terrain height information is obtained by extracting the characteristics of gray values, gradients and the like of the image blocks and estimating the elevation value of the area by combining a support vector machine regression algorithm. And calculating the terrain characteristic parameters such as gradient, slope direction and the like of the area according to the terrain height information. And for the vegetation area, carrying out density classification on vegetation by extracting texture features such as a color histogram, a symbiotic matrix and the like of the image block and combining a random forest classification algorithm to obtain vegetation density information. And according to the vegetation density information, processing the vegetation area by adopting a morphological operation algorithm, and extracting vegetation characteristic parameters such as vegetation coverage and the like. And for the tree area, the tree distribution position information is obtained by extracting the edge characteristics of the image block and detecting the trunk position by combining with a Hough transformation algorithm. And according to the tree distribution position, dividing the tree crowns by adopting a watershed dividing algorithm to obtain a single tree crown area. And estimating the height of the tree by extracting the external rectangle size of the crown area and combining a least square fitting algorithm to obtain the tree height information. And (3) integrating the topographic feature parameters, the vegetation feature parameters and the tree height information, establishing a dead zone judging model by adopting a logistic regression algorithm, and judging the cause of the potential dead zone. The method comprises the steps of judging a dead zone caused by terrain shielding when the gradient is larger than a certain threshold value and the gradient is opposite to a monitoring point, judging the dead zone caused by vegetation shielding when the vegetation coverage is larger than a certain threshold value and the vegetation density is larger, and judging the dead zone caused by tree shielding when the tree height is larger than a certain threshold value and the tree distribution position is closer to the monitoring point. And simulating a visual line from the monitoring point to the blind area by adopting a ray projection algorithm according to the blind area discrimination result to obtain the position of the pixel point affected by the blind area. And matching the position information with the original monitoring image, determining image data points affected by the blind area, and marking the image data points as blind area data.
In one possible implementation manner, the decision tree-based classification algorithm classifies the image block types through the color features, the texture features and the shape features of each image block to determine the environment type of each image block, and specifically includes that the decision tree classification algorithm is adopted to classify the environment type of each image block according to the acquired color features, texture features and shape features of each image block to obtain the environment type of each image block, including topography, vegetation and trees. The color histogram feature of each image block is extracted, the color distribution information of the image block is obtained and is used as one of input features of a decision tree classification algorithm to judge the environment category to which the image block belongs. And calculating the texture characteristics of the image block by adopting the gray level co-occurrence matrix, obtaining texture information of the image block, such as contrast, correlation, energy and the like, and using the texture characteristics as another input characteristic of a decision tree classification algorithm for determining the environment category of the image block. The shape information of the image block is obtained by extracting the shape characteristics of the image block, such as area, perimeter, circularity and the like, and is used as the third type input characteristic of the decision tree classification algorithm for judging the environment category to which the image block belongs. And constructing an environment category classification model by adopting a decision tree classification algorithm according to the color features, the texture features and the shape features of the acquired image blocks, and optimizing decision tree model parameters by training sample data to obtain an optimal decision tree classification model. And (3) carrying out classification prediction on the characteristics of each image block by adopting a constructed decision tree classification model, and judging the environment category of each image block by node division and path selection of the decision tree to obtain the category label of the image block, namely the terrain, vegetation or tree. And integrating the environmental information of the whole image according to the environmental category label of each image block predicted by the decision tree classification model to obtain the environmental category distribution conditions of different areas in the image, and determining the distribution characteristics of the topography, vegetation and trees in the image. And judging the duty ratio and the spatial distribution characteristics of different environment categories in the image through analyzing the classification result of the environment categories of the image block, and acquiring the integral characteristics of the image environment, thereby providing basis for subsequent environment identification and analysis.
In one possible implementation manner, the method comprises the steps of extracting gray values and gradients of each image block in the terrain area, estimating elevation values of the terrain area in combination with a support vector machine regression algorithm to obtain terrain height information, and calculating to obtain terrain feature parameters according to the terrain height information. And carrying out gradient calculation on each image block by adopting a gradient operator, obtaining gradient values of the image blocks in the horizontal and vertical directions, and obtaining gradient feature vectors of each image block by taking the gradient values as gradient features of the image blocks. And taking the gray feature vector and the gradient feature vector of the image block as input, and estimating the elevation value of each image block through a support vector machine regression algorithm to obtain the estimated elevation value corresponding to each image block. And constructing an elevation value matrix of the terrain area according to the position information and the estimated elevation value of each image block to obtain complete terrain elevation information. And calculating the slope, slope direction, curvature and other terrain characteristic parameters of the terrain area based on the terrain height information to obtain the terrain characteristic description of the terrain area. Combining the terrain height information and the terrain characteristic parameters, constructing a three-dimensional model of the terrain area, realizing accurate expression and analysis of the terrain area, and providing data support for subsequent terrain application.
In a possible implementation manner, in the vegetation area, a color histogram and a co-occurrence matrix of each image block are extracted, and a random forest classification algorithm is combined to classify the density of vegetation, so as to obtain vegetation density information, and the vegetation density information is obtained by calculating according to the vegetation density information, which specifically includes:
According to the acquired remote sensing image data, a vegetation index threshold segmentation method is adopted, and pixels with the NDVI larger than a set threshold are divided into vegetation areas by calculating normalized vegetation indexes NDVI, so that a vegetation area mask image is obtained. After the vegetation region mask image is obtained, an image blocking method is adopted, and a sliding window with a fixed size is arranged to slide on the vegetation region image with a certain step length, so that a plurality of image blocks are obtained. According to the extracted image blocks, a color histogram calculation method is adopted, and the histogram vector representing the color characteristics of the image blocks is obtained by counting the color distribution condition of pixels in each image block. According to the extracted image block, a gray level co-occurrence matrix calculation method is adopted, and a co-occurrence matrix representing the texture characteristics of the image block is obtained by counting the co-occurrence condition of pixel gray values in the image block. After the color histogram vector and the gray level co-occurrence matrix are obtained, a feature splicing method is adopted, and two features are spliced according to a certain sequence to obtain a fused image block feature vector. According to the actual vegetation density of the vegetation area, a manual labeling method is adopted, and a representative image block is selected, and is labeled by professionals to obtain a sample data set for training and testing. After the characteristic vectors of the spliced image blocks and the corresponding vegetation density labels are obtained, training and classifying the characteristics of the image blocks by constructing a plurality of decision trees by adopting a random forest classification algorithm to obtain the vegetation density class of each image block. According to the classification of the vegetation density degree of the classified image blocks, a voting strategy is adopted, and the vegetation density distribution condition of the whole vegetation area is obtained by counting the number of the image blocks of each vegetation density level. According to vegetation density distribution conditions, a weighted average method is adopted, corresponding weights are given to image blocks with different density levels, and comprehensive vegetation density information of a vegetation area is obtained through calculation.
In a possible implementation manner, the method comprises the steps of extracting edge characteristics of each image block in a tree area, detecting trunk positions by combining with a Hough transformation algorithm to obtain tree distribution position information, calculating to obtain tree height information according to the tree distribution position information, extracting edge characteristics of trunks in each image block by adopting an edge detection algorithm to obtain trunk edge contour information according to an obtained tree area image, detecting position coordinates of trunks in the image by combining with the trunk edge contour information by using the Hough transformation algorithm to obtain the distribution position information of each tree, calculating height difference between the top position and the bottom position of each tree by adopting a triangulation principle according to the tree distribution position information to obtain the height information of each tree, and generating a tree distribution map comprising the position and the height of each tree by combining with the tree distribution position information according to the obtained height information of each tree to provide data support for subsequent forest management and monitoring.
And S106, deducing and obtaining the monitoring data of the potential monitoring blind area by analyzing the monitoring data of the adjacent area of the potential monitoring blind area, and predicting the long-term change trend of the monitoring data of the potential monitoring blind area by a time sequence prediction algorithm.
In the embodiment, a K-means clustering algorithm is adopted to divide a monitoring area according to the situation of the terrain and vegetation, so that subareas with different characteristics are obtained. And predicting and deducting the data of the blind area by using the existing monitoring data in the subarea through a Kriging spatial interpolation algorithm. When dividing the subareas, optimizing the subarea boundaries through a Voronoi graph algorithm, and avoiding overlapping of monitoring coverage areas. And generating a monitoring equipment configuration rule by adopting a decision tree algorithm aiming at the terrain and vegetation characteristics of different subareas. The sensitivity and sampling frequency parameters of the device are dynamically adjusted according to the rules. And optimizing the decision tree through an AdaBoost algorithm, and improving the adaptability and the robustness of the configuration rule. In the data prediction and deduction process, a self-adaptive Kalman filtering algorithm is adopted to dynamically correct a prediction result, so that a prediction error is reduced. And carrying out long-term trend prediction on the blind area data by combining historical data and seasonal factors through an ARIMA time sequence model. In the deployment and scheduling of the monitoring equipment, an ant colony algorithm is adopted to solve the problems of optimal layout positions and path planning. And the Dijkstra shortest path algorithm is used for optimizing the equipment maintenance and data return routes and reducing the operation and maintenance cost. And finally, carrying out anomaly detection on the monitoring data through a DBSCAN density clustering algorithm, and timely finding and positioning an anomaly region. And dynamically adjusting a monitoring strategy according to the position and the characteristics of the abnormal region, and starting a targeted monitoring task. And evaluating the importance degree of the abnormal region through a PageRank algorithm, and optimizing the allocation and scheduling of the monitoring resources.
In some embodiments, the step S106 specifically includes:
Dividing a monitoring area by adopting a K-means clustering algorithm according to the topography condition and the vegetation condition to obtain subareas with different characteristics, and optimizing the area boundary of the subareas by using a Voronoi graph algorithm;
Determining a plurality of sub-areas adjacent to the potential monitoring blind area, predicting and deducting the monitoring data of the potential monitoring blind area by adopting a Kriging space interpolation algorithm according to the monitoring data in the plurality of sub-areas adjacent to the potential monitoring blind area, and dynamically correcting according to a self-adaptive Kalman filtering algorithm to obtain the monitoring data of the blind area;
And predicting the long-term change trend of the monitoring data of the potential monitoring blind area through an ARIMA time sequence model according to the blind area monitoring data and seasonal factors in a preset time period, and obtaining the predicted monitoring data of the potential monitoring blind area.
In the embodiment, according to the topography condition and vegetation condition of a monitoring area, clustering analysis is carried out on topography and vegetation characteristics of the monitoring area by adopting a K-means clustering algorithm, areas with similar topography and vegetation characteristics are obtained, and subareas with different characteristics are obtained through clustering analysis. Optimizing the region boundaries of the subregions with different characteristics obtained in the previous step by adopting a Voronoi graph algorithm, and obtaining the optimal boundaries of the subregions by adopting the Voronoi graph algorithm to obtain the optimized boundaries of the subregions. And obtaining the topographic features and the vegetation features of each sub-area according to the optimized sub-area boundaries, and determining the topographic type and the vegetation type of each sub-area by analyzing the topographic and vegetation features of each sub-area. According to the terrain type and vegetation type of each sub-area, different monitoring strategies and monitoring methods are adopted to obtain the monitoring data of each sub-area, and the ecological environment condition of each sub-area is judged by analyzing the monitoring data. According to the ecological environment condition of each sub-area, different protection measures and management measures are adopted, and the ecological environment improvement condition of each sub-area is obtained by implementing the protection measures and the management measures, so that the overall ecological environment condition of the monitoring area is determined.
The method comprises the steps of obtaining position information of a potential monitoring blind area according to grids divided by a monitoring area, determining a plurality of sub-areas adjacent to the potential monitoring blind area through analyzing grids around the potential monitoring blind area, obtaining monitoring data in the plurality of sub-areas adjacent to the potential monitoring blind area through a wireless sensor network, obtaining effective monitoring data available for space interpolation through data preprocessing, obtaining preliminary prediction monitoring data of the potential monitoring blind area through constructing a semi-variation function and a covariance matrix according to the principle of a Kriging space interpolation algorithm, obtaining dynamic correction values of the potential monitoring blind area monitoring data through dynamically updating a state equation and an observation equation according to a self-adaptive Kalman filtering algorithm, obtaining final blind area monitoring data of the potential monitoring blind area through weighted fusion according to the potential monitoring blind area preliminary prediction monitoring data obtained by the Kriging space interpolation algorithm, judging the condition of the potential blind area through an environment monitoring visual platform, determining whether the pollution condition of the potential blind area exceeds the pollutant concentration, and accordingly improving the monitoring accuracy of the whole monitoring network is improved.
And collecting and arranging blind area monitoring data in a preset time period to obtain historical monitoring data of potential monitoring blind areas, wherein the historical monitoring data are used as input data of an ARIMA model. Preprocessing the data according to the collected historical monitoring data and combining seasonal factors to remove abnormal values and noise data, so as to obtain a normalized monitoring data sequence. And carrying out stabilization treatment on the normalized monitoring data sequence by adopting a difference method, eliminating the non-stationarity of the data, and obtaining stable time sequence data. And determining the order parameters of the ARIMA model by carrying out autocorrelation and partial correlation analysis on the stabilized time sequence data to obtain an optimal model structure. And estimating model parameters by adopting a maximum likelihood estimation method according to the determined ARIMA model structure to obtain coefficients and residual error items of the model. And (3) judging the fitting effect and the prediction capability of the model by carrying out diagnostic test on the estimated ARIMA model, so as to ensure the effectiveness and the reliability of the model. And according to the ARIMA model passing verification, combining the historical monitoring data of the potential monitoring blind area and seasonal factors, predicting the monitoring data in a period of time in the future, and obtaining predicted monitoring data of the potential monitoring blind area. And analyzing and evaluating the predicted monitoring data to judge the long-term change trend of the potential monitoring blind area, and providing decision basis for optimizing the monitoring blind area.
And S107, monitoring data of each monitoring device, analysis results of the target overlapping area and prediction results of the potential monitoring blind areas are transmitted to a data center through a satellite Internet of things, and an environment monitoring report, a fire risk map and a temperature and humidity change trend chart are generated according to data analysis.
In the embodiment, the data center comprehensively evaluates forest environments including fire early warning, weather changes and ecological changes, generates an environment monitoring report through data analysis, and makes a fire risk map and a temperature and humidity change trend map. And transmitting the processed monitoring data to a data center through the satellite Internet of things, and summarizing and integrating the data of each device. And analyzing the wind direction change by adopting an empirical orthogonal decomposition algorithm according to the analysis result of the overlapping region, acquiring humidity fluctuation data, and evaluating the fire risk by a fuzzy comprehensive evaluation method. The data center comprehensively evaluates forest environments including fire early warning, weather changes and ecological changes. And according to the fire disaster early warning result, carrying out data analysis by adopting a random forest algorithm to obtain an environment monitoring report. And according to the meteorological change data, a fire risk map is manufactured through a support vector machine algorithm. And preparing a temperature and humidity change trend graph by adopting a time sequence analysis method according to the ecological change data. And if the fire risk is higher than the threshold value, triggering fire early warning, and sending early warning information to the monitoring equipment through the satellite communication system. If the humidity is lower than the threshold value and the temperature is higher than the threshold value, judging that the weather is high in fire risk, and improving the monitoring frequency. And planning a patrol route by adopting an ant colony algorithm according to the fire risk map, and optimizing the patrol efficiency. And predicting weather changes of a week in the future through a hidden Markov model according to the temperature and humidity change trend graph, and adjusting a monitoring strategy. And acquiring real-time data of each monitoring device, and comprehensively analyzing through a data fusion algorithm to judge whether an abnormal condition exists. If the abnormal situation is found, an emergency plan is started, and rescue force is scheduled.
It can be understood that a micro weather meter or weather sensor can be deployed at the key position of each monitoring device position or the nearby position, and is used for collecting weather data such as temperature, humidity, atmospheric pressure, wind speed, wind direction, PM2.5, PM10, noise, illuminance, rainfall and the like in real time, and a multi-level monitoring system is built together with the monitoring device for smoke and fire identification by combining with the satellite internet of things terminal.
In some embodiments, the monitoring data includes wind speed and direction data, temperature monitoring data, and humidity monitoring data, the method further comprising:
Analyzing the degree of difference between the wind speed and direction data of each monitoring device in the target overlapping area by determining the influence of the topographic feature data on the wind speed and direction data;
according to the temperature change trend in the forest environment, analyzing the degree of difference between the temperature monitoring data of each monitoring device in the target overlapping area;
and determining the humidity characteristics of the forest environment according to the vegetation type spatial distribution map and the water source spatial distribution map in the forest environment, and analyzing the degree of variability among humidity monitoring data of each monitoring device in the target overlapping area through the humidity characteristics of the forest environment.
In the embodiment, a multiple regression model is established according to the topographic feature data of the forest environment and the wind speed and direction data of the forest environment to obtain influences of the topographic feature on the wind speed and direction, different topographic type data, the wind speed data and the wind direction data of the forest environment are adopted to conduct variance analysis to obtain wind speed and direction differences among different topographic types, a temperature time sequence chart is drawn according to the temperature time sequence data of the forest environment, the stability of the temperature data is judged, the ARIMA model order is determined according to an autocorrelation and bias autocorrelation chart, the ARIMA model is established, future temperatures are predicted through the ARIMA model to obtain temperature prediction results and confidence intervals, temperature prediction result differences of different devices are compared, a vegetation type spatial distribution map and a water source spatial distribution map of the forest environment are obtained, the vegetation type spatial distribution map, the water source spatial distribution map and the humidity data of the forest environment are subjected to superposition analysis to obtain spatial correlation of the humidity and the vegetation type, a spatial interpolation method is adopted to conduct spatial interpolation method to the humidity spatial distribution map, and the humidity spatial distribution map is combined with the water source spatial distribution map, and the spatial distribution map of the humidity and the spatial distribution map is analyzed.
Referring to fig. 2, an embodiment of the present invention provides a forest fire early warning system 2 based on the satellite internet of things, where the system 2 specifically includes:
the first early warning module 201 is configured to establish a three-dimensional space model according to pose data and forest topography data of each monitoring device, simulate a monitoring area of each monitoring device in the three-dimensional space model through a topography map and a satellite image, and determine a potential monitoring blind area and a potential overlapping area in the monitoring area;
a second early warning module 202, configured to determine, based on the three-dimensional spatial model, a plurality of first overlapping regions from a plurality of the potential overlapping regions according to a spatial geometry algorithm, and determine a plurality of target overlapping regions from a plurality of the first overlapping regions according to terrain relief data and tree shielding data;
The third early warning module 203 is configured to obtain monitoring data collected by each monitoring device in the target overlapping area, compare and analyze the monitoring data of each monitoring device according to a correlation coefficient algorithm and a local anomaly factor algorithm, determine consistency and variability of the monitoring data of each monitoring device, and determine a data interference factor according to the consistency and the variability;
The fourth early warning module 204 is configured to perform fusion processing on the monitoring data of different monitoring devices in each target overlapping area according to a data fusion algorithm, so as to generate a data fusion model, where the data fusion model is used to eliminate overlapping effects caused between the monitoring data of different monitoring devices in the target overlapping area;
A fifth early warning module 205, configured to analyze a monitored image in the potential monitoring blind area by using an image recognition technology, extract topographic feature data, vegetation density data and tree height data in the monitored image, and determine a blind area cause according to the topographic feature data, the vegetation density data and the tree height data;
A sixth early warning module 206, configured to derive the monitored data of the potential monitored blind area by analyzing the monitored data of the adjacent area of the potential monitored blind area, and predict a long-term variation trend of the monitored data of the potential monitored blind area by a time sequence prediction algorithm;
and the seventh early warning module 207 is configured to transmit the monitoring data of each monitoring device, the analysis result of the target overlapping area, and the prediction result of the potential monitoring blind area to a data center through the satellite internet of things, and generate an environmental monitoring report, a fire risk map, and a temperature and humidity change trend chart according to data analysis.
It can be understood that the content in the embodiment of the forest fire early warning method based on the satellite internet of things shown in fig. 1 is applicable to the embodiment of the forest fire early warning system based on the satellite internet of things, the functions of the embodiment of the forest fire early warning system based on the satellite internet of things are the same as those of the embodiment of the forest fire early warning method based on the satellite internet of things shown in fig. 1, and the beneficial effects achieved by the embodiment of the forest fire early warning method based on the satellite internet of things shown in fig. 1 are the same as those achieved by the embodiment of the forest fire early warning method based on the satellite internet of things shown in fig. 1.
It should be noted that, because the content of information interaction and execution process between the above systems is based on the same concept as the method embodiment of the present invention, specific functions and technical effects thereof may be referred to in the method embodiment section, and will not be described herein.
It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of the functional units and modules is illustrated, and in practical application, the above-described functional distribution may be performed by different functional units and modules according to needs, i.e. the internal structure of the system is divided into different functional units or modules to perform all or part of the above-described functions. The functional units and modules in the embodiment may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit, where the integrated units may be implemented in a form of hardware or a form of a software functional unit. In addition, the specific names of the functional units and modules are only for distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working process of the units and modules in the above system may refer to the corresponding process in the foregoing method embodiment, which is not described herein again.
Referring to fig. 3, the embodiment of the invention further provides a computer device 3, which comprises a memory 302, a processor 301 and a computer program 303 stored on the memory 302, wherein the computer program 303, when executed on the processor 301, implements the forest fire early warning method based on the satellite internet of things according to any one of the above methods.
The computer device 3 may be a desktop computer, a notebook computer, a palm computer, a cloud server, or the like. The computer device 3 may include, but is not limited to, a processor 301, a memory 302. It will be appreciated by those skilled in the art that fig. 3 is merely an example of the computer device 3 and is not meant to be limiting as the computer device 3, and may include more or fewer components than shown, or may combine certain components, or different components, such as may also include input-output devices, network access devices, etc.
The Processor 301 may be a central processing unit (Central Processing Unit, CPU), the Processor 301 may also be other general purpose processors, digital signal processors (DIGITAL SIGNAL processors, DSP), application SPECIFIC INTEGRATED Circuit (ASIC), off-the-shelf Programmable gate array (Field-Programmable GATE ARRAY, FPGA) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory 302 may in some embodiments be an internal storage unit of the computer device 3, such as a hard disk or a memory of the computer device 3. The memory 302 may also be an external storage device of the computer device 3 in other embodiments, such as a plug-in hard disk, a smart memory card (SMART MEDIA CARD, SMC), a Secure Digital (SD) card, a flash memory card (FLASH CARD) or the like, which are provided on the computer device 3. Further, the memory 302 may also include both an internal storage unit and an external storage device of the computer device 3. The memory 302 is used to store an operating system, application programs, boot loader (BootLoader), data, and other programs, such as program code for the computer program. The memory 302 may also be used to temporarily store data that has been output or is to be output.
The embodiment of the invention also provides a computer readable storage medium, on which a computer program is stored, which when being run by a processor, realizes the forest fire early warning method based on the satellite Internet of things according to any one of the methods.
In this embodiment, the integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the present application may implement all or part of the flow of the method of the above embodiments, and may be implemented by a computer program to instruct related hardware, where the computer program may be stored in a computer readable storage medium, and when the computer program is executed by a processor, the computer program may implement the steps of each of the method embodiments described above. Wherein the computer program comprises computer program code which may be in source code form, object code form, executable file or some intermediate form etc. The computer readable medium may include at least any entity or device capable of carrying computer program code to a camera device/terminal equipment, a recording medium, a computer Memory, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), an electrical carrier signal, a telecommunications signal, and a software distribution medium. Such as a U-disk, removable hard disk, magnetic or optical disk, etc. In some jurisdictions, computer readable media may not be electrical carrier signals and telecommunications signals in accordance with legislation and patent practice.
In the foregoing embodiments, the descriptions of the embodiments are emphasized, and in part, not described or illustrated in any particular embodiment, reference is made to the related descriptions of other embodiments.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
In the disclosed embodiments of the application, it should be understood that the disclosed apparatus/terminal device and method may be implemented in other ways. For example, the apparatus/terminal device embodiments described above are merely illustrative, e.g., the division of the modules or units is merely a logical function division, and there may be additional divisions in actual implementation, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection via interfaces, devices or units, which may be in electrical, mechanical or other forms.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.