wherein, (x'₀,y′₀) Is the initial position of the child, d is the step length of the child, theta is the course angle, (x'_k,y′_k) The position of the child at step k.

The step length of the infant can be judged to be obtained according to the height of the infant and the acceleration collected by an accelerometer in the inertial navigation equipment; the course angle judges which direction the infant takes to according to a gyroscope and a magnetometer in the inertial navigation equipment; the PC can also determine when the child takes one step by receiving the acceleration collected by the accelerometer.

In the foregoing method for monitoring activity of an infant area, in the step C, the operation of stitching the unmatched visual motion trajectory and inertial motion trajectory to obtain the visual-inertial motion trajectory of the infant includes:

extracting the time t of each step of the inertia motion track_i；

At the judgment t_iPosition (x ″) of the visual motion trajectory at the moment_i,y″_i) If present, the position (x) of the visual-inertial motion trajectory is determined_i,y_i)＝(x″_i,y″_i)；

At the judgment t_i+1Position (x ″) of the visual motion trajectory at the moment_i+1,y″_i+1) When the inertial motion track does not exist, the position (x ') of the corresponding time step of the inertial motion track is taken'_k,y′_k) And with t_iTime t and_i+n+1the position coordinate point where the time visual motion trail exists is corrected and calculated to obtain a corrected position (x ″)'_k,y″′_k) And assigning the position of the corresponding moment of the visual-inertial motion track, and calculating the correction calculation by the following formula:

wherein, (x'_k,y″′_k) After being corrected for the inertia motion trackThe location of the step; (x ″)_i,y″_i) Is t_iPosition coordinate point of moment visual motion trail, (x ″)_i+n+1,y″_i+n+1) Is t_i+n+1Position coordinate points of moment visual motion trail, (x'_i+1,y′_i+1) Is t_i+1Coordinates of steps of the moment inertial motion trajectory; (x'_k,y′_k) Coordinates of corresponding time steps of the inertia motion track are obtained; (x'_i+n+1,y′_i+n+1) Is t_i+n+1Coordinates of steps of the moment inertial motion trajectory. The inertial motion trail and the visual motion trail are combined to form a visual-inertial motion trail, the visual-inertial motion trail effectively overcomes the defects of the inertial motion trail and the visual motion trail, and the two motion trails are combined to realize complementation, so that the obtained infant motion trail has higher accuracy.

In the foregoing method for monitoring baby area activity, in step D, the operation of evaluating the personality preference of the baby according to the feature value to generate a personality preference report for each baby includes:

evaluating the concentration degree of the infant according to the characteristic value;

evaluating the liveness of the infant according to the characteristic value;

evaluating the preference of the infant to each subspace according to the characteristic values;

evaluating the attraction of each subspace according to the characteristic value;

and summarizing the concentration degree, the activity degree, the preference degree of each subspace and the attraction of each subspace of the infant into a personality preference report.

The infant activity management system collects the concentration degree, the activity degree, the preference degree of each subspace and the attraction of each subspace of an infant into a personality preference report, can accurately feed back the regional activity condition of the infant to each parent, and is convenient for the parents to know the personality preference of the child.

In the baby area activity monitoring method, in the step D, the operation of evaluating the concentration of the baby according to the feature value includes:

extracting time values of the infants entering or leaving the entrance and exit of the subspace from the visual-inertial motion trail according to the coordinate point information of the entrance and exit of each subspace;

subtracting the time values of the entrance and the exit of the subspace to calculate the staying time T of the infant in each subspace_Kid；

Will stay for a time period T_KidRecording the name, the name and the date of the subspace, the time value for entering the subspace and the stay time into a table established by the name, the name and the date of the subspace of the infant;

all stay time lengths T of the same infant in the table_KidComparing, and recording the maximum value as the maximum single stay time T of the infant_KidMaxAnd further according to the dwell time T_KidMaxTo determine the concentration of the child.

The judgment of good or bad attention can be made by the length T of the stay time_KidMaxThe length of time of being absorbed in that should reach with this age bracket infant compares and judges, and the infant is absorbed in the definite ability and can help teacher and the head of a family more comprehensive and more timely discovery problem to the infant is absorbed in the degree to make and guide the correction.

In the baby area activity monitoring method described above, in the step D, the operation of evaluating the liveness of the baby according to the feature value includes:

segmenting an acceleration array consisting of time parameters and acceleration values acquired and recorded by an accelerometer in inertial navigation equipment according to time values of entering and leaving a certain subspace, extracting the acceleration array of the infant in the certain subspace, and averaging the acceleration values in the acceleration array of the certain subspace to obtain an average value of the single-region acceleration of the infant;

and carrying out weighted average on the infant single-region acceleration average values of different subspaces to obtain an infant acceleration average value, and further determining the activity of the infant according to the infant acceleration average value.

Can reflect which infant is better moving according to the infant acceleration average value of difference, which infant is more happy and quiet, can help new teacher to know the infant fast, also can know the liveness of same infant in different periods through infant acceleration average value moreover, can reflect whether the infant is uncomfortable or not the mood is not good, and help the teacher to know the infant better.

In the foregoing method for monitoring baby area activity, in the step D, the operation of evaluating the preference of the baby for each subspace according to the feature value includes:

extracting time values of the infants entering or leaving the entrance and exit of each subspace from the visual-inertial motion trail according to the coordinate point information of the entrance and exit of each subspace;

The stay time T of the same child in the same subspace_KidAccumulating to obtain the accumulated stay time T of the single area of the infant_KidArea；

Accumulating the stay time T of the infant single areas in different subspaces_KidAreaThe comparison is made to confirm the child's preference for each subspace.

The evaluation of the preference of the infant can help teachers and parents to better understand the infant, and can make a targeted scheme to promote personalized development and comprehensive development of the infant.

In the baby area activity monitoring method, in the step D, the operation of evaluating the attraction of each subspace according to the feature value includes:

extracting time values of the children entering or leaving the entrance and the exit of each subspace from the visual-inertial motion trail according to the coordinate point information of the entrance and the exit of each subspace;

subtracting the time values of the entrance and the exit of the subspace to obtain the staying time T of each infant in each subspace_Kid；

The stay time T of all the infants in the same subspace_KidAdding and calculating to obtain the total residence time T of the whole class list area_ClassAreaSum；

Accumulating the stay time T of the whole class list area_ClassAreaSumDividing the total activity duration of the subspace by the upper limit number of people preset in the subspace to obtain the full seat rate of the subspace;

accumulating the stay time T of the whole class list area_ClassAreaSumDividing the total number of the infants entering the subspace to obtain the average staying time T of the whole class single region_{ClassAreaAver}；

Judging the subspace with the full rate of less than 90 percent of the subspace as a non-hot area and accumulating the stay time T according to the whole class list area_ClassAreasumSorting to obtain an attraction sorting table of a non-hot area;

judging the subspace with the full rate of the subspace being more than or equal to 90 percent as a hot area and according to the average stay time T of the whole class single area_{ClassAreaAver}Sorting to obtain an attraction sorting table of the hot area;

and confirming the attraction of each subspace to the infant through the attraction sequencing table of the non-hot area and the attraction sequencing table of the hot area.

The evaluation of the attractiveness of the subspace may help teachers evaluate and adjust the suitability of the regional material difficulty.

Compared with the prior art, the method for monitoring the regional activity of the infant has the following advantages:

1. according to the invention, the activity data of the whole class of infants in the regional activity room can be collected and analyzed in real time without requiring infant operation and facing the screen, and the personality and preference form of each infant can be generated through the evaluation of the infant activity, the evaluation of the infant concentration, the evaluation of the infant preference and the evaluation of the regional attraction.

2. According to the invention, through the evaluation of the liveness of the infant and further through the liveness conditions of the infant in different periods, the mood or the physical health condition of the infant can be known in time, and teachers and parents can be helped to pay attention to the physical and mental health of the infant in time.

Drawings

Fig. 1 is a control flow chart of the present invention.

Detailed Description

The following are specific embodiments of the present invention and are further described with reference to the drawings, but the present invention is not limited to these embodiments.

Before evaluating the personality and preference of the infant through the infant regional activity monitoring method and accurately helping teachers and parents to know the infant deeply, a regional activity room is arranged firstly, the regional activity room is composed of a plurality of subspaces, each subspace is provided with only one entrance, boundary geographical labels are pasted on each subspace to form independent spaces, each boundary geographical label is composed of a node and a straight line passing through the node, and the distance between the adjacent boundary geographical label nodes can be set to be 20 cm. And pasting an entrance geographic label near the entrance of the subspace. The entrance and exit geographic label is a two-dimensional code containing the area name, and each subspace is provided with a pair of entrance and exit geographic labels.

And then arranging acquisition equipment, respectively placing one camera equipment at four vertex angles of a classroom, transmitting signals acquired by the camera equipment to a PC (personal computer) through a network cable, carrying an inertial navigation equipment by each infant, wherein the acquired signals are transmitted to the PC through Bluetooth.

After the acquisition equipment is set, the acquisition equipment can be set to start working when the infant starts regional activities and close working after the infant finishes regional activities, so that energy is saved. As shown in fig. 1, after the work is started, image data of the movement of the infant in the regional activity room is acquired by the camera device, wherein the image data of the regional activity room comprises the image data of the regional activity room and the image data of the infant in the regional activity room, the acquisition frequency of the camera device is 60 frames/second, and the movement speed of the infant is less than 6 m/second, so that the distance between two adjacent frames of the infant is generally less than 0.1 m; the inertial navigation equipment worn by each infant collects the motion data of each infant, wherein the motion data comprises acceleration and a course angle, and the inertial navigation equipment can set acceleration sampling frequency, such as sampling frequency of 10 times/second or other numerical values; the camera equipment sends the collected image data to the PC at the image sampling frequency, and the inertial navigation equipment sends the collected motion data and the time of collecting the motion data to the PC at the acceleration sampling frequency, so that the accuracy of data judgment in the later period is ensured;

the PC collects image data collected by each camera device and analyzes the collected image data, firstly, the boundary geographical labels and the entrance geographical labels are identified through an image identification technology, and N pairs of connection lines of the entrance geographical labels and a plurality of boundary geographical label nodes can be obtained, wherein N is the number of the regions. And then connecting the entrance and exit geographic labels with the same name, connecting nodes at two ends of the boundary geographic label with other nearest boundary geographic label nodes or entrance and exit geographic labels, and connecting the unconnected entrance and exit geographic labels with the nearest boundary geographic label nodes. Thus, N closed connecting lines containing the region inlets and outlets can be obtained, the space surrounded by the connecting lines is the range of the corresponding region, the region activity room can be divided into N +1 subspaces, wherein the subspaces comprise N regions and a public space, and different regions are connected with the public space through the inlets and outlets; given that the angle between adjacent walls in a classroom is 90, the elevation of all boundary geotag nodes is substantially uniform, and thus the azimuth of all boundary geotags can be estimated. Additionally, given the 20 cm spacing between adjacent boundary geotag nodes, the coordinates of each node can be estimated in conjunction with the previous azimuth. Therefore, a geographical label space model which consists of N +1 subspaces and comprises coordinate nodes can be established; in addition, the minimum distance between the entrances and the exits of different areas is required to be greater than the measurement accuracy of the visual-inertial positioning system, the distance between the entrances and the exits of different subspaces is set to be greater than or equal to 2 meters, and in the process of establishing the model, if the distance between the entrances and the exits of different areas is less than 2 meters, a warning is given out to require the entrance and the exit of the areas to be rearranged.

After the geographical label space model is established, calculating coordinate points of the infant in the geographical label space model, specifically: extracting the infants in the image data by using an SIFT algorithm, judging the subspace of each identified infant, comparing the infants with boundary geographical label nodes in the subspaces when the infants are in the subspaces, and calculating coordinate points of the infants in a geographical label space model; if the child is located near the boundary of the plurality of subspaces, judging which subspace the child is located in according to the previous visual motion trail of the child, and calculating the coordinates of the child; when the child cannot be judged to be in which subspace through the previous visual motion trail of the child, comparing the child with boundary geographical label nodes in each subspace, calculating a coordinate once, and averaging a plurality of coordinate results to obtain a coordinate point of the child in a geographical label spatial model;

then connecting a plurality of coordinate points of the same infant according to the sequence of the time nodes so as to form the visual motion trail of the infant; the operation of the connection also includes the following situations: if the total movement of a certain 'child' is less than two pixels within 1 minute, the motion track point is removed because the distance between two adjacent frames of the child is less than 0.1 meter, and if the coordinate point of the child cannot be found within 0.1 meter of the next frame by a certain motion track, the coordinate point of the child within the next a/10 meter is searched by frame skipping. Where a is the number of skipped frames and a ≦ 10. If the coordinate point of the infant is found through the frame skipping, the images of the infant before and after the frame skipping are compared. If the similarity is high, the baby is considered to be the same baby, otherwise, the movement track is interrupted.

The motion data collected by each inertial navigation device is collected through a PC and is analyzed and processed, and the method specifically comprises the following steps: the acquired motion data is calculated through a PDR algorithm, the step length, the course angle and the time when the baby takes one step are obtained, and then the position of each step of the baby moving in the regional activity room is obtained according to the following formula:

wherein, (x'₀,y′₀) Is the initial position of the child, d is the step length of the child, theta is the course angle, (x'_k,y′_k) The position of the child at step k. Finally, connecting the positions of each step to form the inertial motion track of each infant;

assuming that n infants exist, n continuous inertial motion tracks and a plurality of visual motion tracks can be obtained, then the inertial motion track and the plurality of visual motion tracks of each infant are compared, and m pairs of inertial motion tracks and visual motion tracks with high correlation can be obtained, wherein m is less than or equal to n. If n pairs of inertia motion tracks and vision motion tracks with high correlation are obtained in total, making the inertia-vision motion tracks equal to the vision motion tracks, and correspondingly obtaining the inertia-vision motion tracks of each infant according to the inertia motion tracks;

if m pairs of inertial motion tracks and visual motion tracks with high correlation are obtained in total, and m is smaller than n, splicing the remaining visual motion tracks and the remaining inertial motion tracks until inertial-visual motion tracks of n infants are generated, wherein the specific operation of splicing the visual motion tracks and the inertial motion tracks is as follows:

extracting the time t of each step of the inertia motion track_i；

At the judgment t_iPosition (x ″) of the visual motion trajectory at the moment_i,y″_i) If present, the position (x) of the visual-inertial motion trajectory is determined_i,y_i)＝(x″_i,y″_i) (ii) a Sequentially judging the time of each step of the inertia motion track, and judging t_i+n+1Position (x ″) of the visual motion trajectory at the moment_i+n+1,y″_i+n+1) If present, the position (x) of the visual-inertial motion trajectory is determined_i+n+1,y_i+n+1)＝(x″_i+n+1,y″_i+n+1) (ii) a In this judgment, t is found_i+1Time to t_i+nThe position of the visual motion track does not exist at the moment, the inertial motion track is analyzed, and the inertial motion track t is extracted_i+1Time to t_i+nStep (x ') corresponding to time'_k,y′_k) Let a t_i+1Step of moment inertial motion track is (x'_i+1,y′_i+1) Will be t_i+1The step of the time being the position (x ″) where the visual movement trace exists_i,y″_i) Correction calculation is performed to correct the position (x ″) of the step after'_i+1,y″′_i+1) Assigning a position (x) to the visual-inertial motion trajectory_i+1,y_i+1). The correction calculation is calculated by the following formula:

wherein, t_i+1Coordinates (x ″ 'of step corrected by moment inertial motion trajectory'_i+1,y″′_i+1) Then the coordinate is assigned to the visual-inertial motion track t_i+1Position of time (x)_i+1,y_i+1) The above step (1); (x ″)_i,y″_i) Is t_iPosition coordinate point of moment visual motion trail, (x ″)_i+n+1,y″_i+n+1) Is t_i+n+1Position coordinate points of moment visual motion trail, (x'_i+1,y′_i+1) Is t_i+1Coordinates of steps of the moment inertial motion trajectory; (x'_i,y′_i) Is t_iCoordinates of steps of the moment inertial motion trajectory; (x'_i+n+1,y′_i+n+1) Is t_i+n+1Coordinates of steps of the moment inertial motion trajectory. At the judgment t_i+1Time to t_i+nAnd when the position coordinate point of the visual motion track at the moment does not exist, the step correction corresponding to the inertial motion track is taken by repeating the above mode, and then the coordinate value of the corrected step is assigned to the coordinate point of the visual-inertial motion track at the corresponding moment. Thereby, the visual-inertial motion track of each infant is obtained.

And finally, extracting characteristic values from the visual-inertial motion trail of the infant, wherein the characteristic values comprise time values of the infant entering and exiting each subspace of the activity room of the area and the acceleration of the activity in each subspace of the room. The time value of the infant entering or leaving each area of the regional activity room can be extracted and obtained from the visual-inertial motion trail according to the coordinate point information of the entrance and the exit of each area; and then evaluating the personality preference of the infant according to the characteristic values, wherein the personality preference comprises concentration degree, liveness degree, subspace preference degree, subspace attraction and the like. Wherein, the operation of concentration degree evaluation is as follows:

subtracting the time values of the entrance and the exit of the subspace to calculate the staying time T of the infant in each subspace_Kid(ii) a Will stay for a time period T_KidRecording the name, the name and the date of the subspace, the time value for entering the subspace and the stay time into a table established by the name, the name and the date of the subspace of the infant; all stay time lengths T of the same infant in the table_KidComparing, and recording the maximum value as the maximum single stay time T of the infant_KidMaxAnd further according to the dwell time T_KidMaxTo determine the concentration of the child. Such as the length of stay T_KidMaxComparing the time with the concentration time of the infant in the age group, and judging the time T of stay_KidMaxIf the length of time that the infant should reach is longer than the age, the attentiveness of the infant is good, otherwise, the attentiveness of the infant is poor.

The activity evaluation operation comprises the following steps:

and segmenting an acceleration array consisting of time parameters and acceleration values acquired and recorded by an accelerometer in the inertial navigation equipment according to the time values of entering and leaving a certain subspace, wherein the acceleration array is an array with a first column of time, other columns of acceleration values and subspace names, and the number of rows equal to the sampling frequency multiplied by the sampling time length. Extracting acceleration arrays of the infants in the same subspace, and averaging acceleration values in the acceleration arrays of the same subspace to obtain an average value of the single-region acceleration of the infants;

and carrying out weighted average on the infant single-region acceleration average values of different subspaces to obtain an infant acceleration average value, and further determining the activity of the infant according to the infant acceleration average value. The acceleration average value of each infant is stored in a table, and a teacher can know which infants are better and which infants are more happy and calm according to different infant acceleration average values, so that the method can help a new teacher to quickly know the infants.

The operation of evaluating the preference degree of the children on each subspace comprises the following steps:

subtracting the time values of the entrance and the exit of the subspace to calculate the staying time T of the infant in each subspace_Kid(ii) a The stay time T of the same child in the same subspace_KidAccumulating to obtain the accumulated stay time T of the single area of the infant_KidArea(ii) a Accumulating the stay time T of the infant single areas in different subspaces_KidAreaThe comparison is made to confirm the child's preference for each subspace. Dwell time T_KidAreaThe larger the size, the greater the child's preference for that subspace.

The operation of the attraction evaluation of each subspace includes:

subtracting the time values of the entrance and the exit of the subspace to obtain the staying time T of each infant in each subspace_Kid(ii) a The stay time T of all the infants in the same subspace_KidAdding and calculating to obtain the total residence time T of the whole class list area_ClassAreasum(ii) a Accumulating the stay time T of the whole class list area_ClassAreaSumDividing the total activity duration of the subspace by the upper limit number of people preset in the subspace to obtain the full seat rate of the subspace; accumulating the stay time T of the whole class list area_ClassAreaSumDividing the total number of the infants entering the subspace to obtain the average staying time T of the whole class single region_{ClassAreaAver}(ii) a Judging the subspace with the full rate of less than 90 percent of the subspace as a non-hot area and accumulating the stay time T according to the whole class list area_ClassAreaSumSorting to obtain an attraction sorting table of a non-hot area; judging the subspace with the full rate of the subspace being more than or equal to 90 percent as a hot area and according to the average stay time T of the whole class single area_{ClassAreaAver}Sorting to obtain an attraction sorting table of the hot area; and confirming the attraction of each subspace to the infant through the attraction sequencing table of the non-hot area and the attraction sequencing table of the hot area.

Finally, the concentration degree, the liveness degree and the preference degree of the infants are made into a personality and preference report of each infant and are posted in the individual growth file of the infant, so that the regional activity condition of the infant can be accurately fed back to each parent, and the personalized development and the comprehensive development of the infant are promoted.

The specific embodiments described herein are merely illustrative of the spirit of the invention. Various modifications or additions may be made to the described embodiments or alternatives may be employed by those skilled in the art without departing from the spirit or ambit of the invention as defined in the appended claims.

Claims

Translated fromChinese

1.一种幼儿区域活动监测方法，其特征在于，包括如下步骤：1. a kind of young child area activity monitoring method, is characterized in that, comprises the steps:

A、通过摄像设备采集幼儿在区域活动室活动的图像数据，由每个幼儿身上佩戴的惯性导航设备采集每个幼儿的运动数据；A. The image data of the children's activities in the regional activity room is collected through the camera equipment, and the motion data of each child is collected by the inertial navigation equipment worn by each child;

B、通过PC机收集摄像设备采集的图像数据并进行分析处理，获得若干个视觉运动轨迹，通过PC机收集各惯性导航设备采集的运动数据并进行分析处理，获得每个幼儿的惯性运动轨迹；B. Collect and analyze the image data collected by the camera equipment through the PC to obtain several visual motion trajectories, collect the motion data collected by the inertial navigation equipment through the PC and analyze and process to obtain the inertial motion trajectory of each child;

D、从幼儿的视觉-惯性运动轨迹中提取特征值，根据特征值评估幼儿的性格喜好从而生成每个幼儿的性格喜好报告，所述特征值包括幼儿出入室内各个活动区域的时间值以及在室内各个活动区域活动的加速度。D. Extract characteristic values from the visual-inertial motion trajectory of the child, and evaluate the child's character preference according to the characteristic value to generate a report of each child's character preference. The characteristic value includes the time value of the child entering and leaving each activity area in the room and the Acceleration of activity in each activity area.

2.根据权利要求1所述的幼儿区域活动监测方法，其特征在于，在所述步骤B中，获得若干个视觉运动轨迹的操作包括：2. The method for monitoring activity in the area of young children according to claim 1, wherein, in the step B, the operation of obtaining several visual motion trajectories comprises:

根据采集的图像数据识别出区域活动室的分界线地理标签和出入口地理标签，从而建立由若干个子空间组成且包含坐标节点的地理标签空间模型；According to the collected image data, the boundary line geotags and the entrance and exit geotags of the regional activity room are identified, so as to establish a geotag space model composed of several subspaces and including coordinate nodes;

根据采集的图像数据识别出区域活动室内的幼儿及幼儿所处的子空间，进而计算出幼儿在地理标签空间模型中的坐标点；According to the collected image data, the children in the regional activity room and the subspace where the children are located are identified, and then the coordinate points of the children in the geotag space model are calculated;

将同一幼儿的多个坐标点根据时间节点的先后顺序进行连接从而形成幼儿的视觉运动轨迹。The multiple coordinate points of the same child are connected according to the sequence of time nodes to form the child's visual movement track.

3.根据权利要求2所述的幼儿区域活动监测方法，其特征在于，在所述步骤B中，计算出幼儿在地理标签空间模型中的坐标点的操作包括：3. The method for monitoring children's regional activities according to claim 2, wherein, in the step B, the operation of calculating the coordinates of the children in the geographic tag space model comprises:

使用SIFT算法提取图像数据中的幼儿，判断每个识别出来的幼儿是处于哪个子空间内，在幼儿处于子空间内时，将幼儿与该子空间中的分界线地理标签节点进行对比，计算幼儿在地理标签空间模型中的坐标点；Use the SIFT algorithm to extract the children in the image data, and determine which subspace each identified child is in. When the child is in the subspace, compare the child with the boundary geotag nodes in the subspace, and calculate the child Coordinate points in the geotagged spatial model;

在幼儿处在多个子空间的界限附近时，通过该幼儿之前的视觉运动轨迹判断其处于哪个子空间内，从而计算幼儿的坐标；在通过幼儿之前的视觉运动轨迹无法判断幼儿处于哪个子空间内时，将幼儿和每一个子空间中的分界线地理标签节点都进行对比并都计算一次坐标，然后将多个坐标结果进行平均化处理从而获得该幼儿在地理标签空间模型中的坐标点。When a child is near the boundaries of multiple subspaces, the child's previous visual movement trajectory is used to determine which subspace it is in, so as to calculate the child's coordinates; it is impossible to determine which subspace the child is in through the child's previous visual movement trajectory When , compare the child with the boundary line geotag nodes in each subspace and calculate the coordinates once, and then average the multiple coordinate results to obtain the coordinate point of the child in the geotag space model.

4.根据权利要求1或2或3所述的幼儿区域活动监测方法，其特征在于，在所述步骤B中，获得每个幼儿的惯性运动轨迹的操作包括：4. The method for monitoring the activity of a child's area according to claim 1, 2 or 3, wherein, in the step B, the operation of obtaining the inertial motion trajectory of each child comprises:

通过PDR算法对采集的运动数据进行推算从而获得幼儿在区域活动室内活动的每一步位置，再将每一步位置进行连接从而形成每个幼儿的惯性运动轨迹，每一步位置由以下公式获得：The collected motion data is calculated by the PDR algorithm to obtain the position of each step of the child's activities in the regional activity room, and then the position of each step is connected to form the inertial motion trajectory of each child. The position of each step is obtained by the following formula:

其中，(x′₀,y′₀)为幼儿的初始位置，d为幼儿的步长，θ为航向角，(x′_k,y′_k)为幼儿的第k步的位置。Among them, (x′₀ , y′₀ ) is the initial position of the infant, d is the step length of the infant, θ is the heading angle, and (x′_k , y′_k ) is the position of the infant’s kth step.

5.根据权利要求1或2或3所述的幼儿区域活动监测方法，其特征在于，在所述步骤C中，将不匹配的视觉运动轨迹与惯性运动轨迹进行拼接从而获得幼儿的视觉-惯性运动轨迹的操作包括：5. according to claim 1 or 2 or 3 described children's area activity monitoring method, it is characterized in that, in described step C, by unmatched visual motion track and inertia motion track are spliced so as to obtain the visual-inertial of young child The operations of the motion track include:

提取惯性运动轨迹每一步的时间t_i；Extract the time t_i of each step of the inertial motion trajectory;

在判断t_i时刻视觉运动轨迹的位置(x″_i,y″_i)存在时，则令视觉-惯性运动轨迹的位置(x_i，y_i)＝(x″_i,y″_i)；When judging the existence of the position (x″_i , y″_i ) of the visual motion trajectory at time t_i , let the position (x_i , y_i ) of the visual-inertial motion trajectory = (x″_i , y″_i );

在判断t_i+1时刻视觉运动轨迹的位置(x″_i+1,y″_i+1)不存在时，取惯性运动轨迹相应时刻步的位置(x′_k,y′_k)并以t_i时刻和t_i+n+1时刻视觉运动轨迹存在的位置坐标点进行修正计算，从而获得修正后的位置(x″′_k,y″′_k)并赋值给视觉-惯性运动轨迹相应时刻的位置，修正计算由以下公式进行计算：When judging that the position (x″_i+1 , y″_i+1 ) of the visual motion trajectory does not exist at time t_i+1 , take the position (x′_k , y′_k ) of the corresponding moment step of the inertial motion trajectory and use t Correction and calculation are performed on the position coordinate points existing in the visual motion trajectory at time_i and time t_i+n+1 , so as to obtain the corrected position (x″′_k , y″′_k ) and assign it to the corresponding moment of the visual-inertial motion trajectory The position, correction calculation is calculated by the following formula:

其中，(x″′_k,y″′_k)为惯性运动轨迹修正后步的位置；(x″_i,y″_i)为t_i时刻视觉运动轨迹的位置坐标点，(x″_i+n+1,y″_i+n+1)为t_i+n+1时刻视觉运动轨迹的位置坐标点，(x′_i+1,y′_i+1)为t_i+1时刻惯性运动轨迹的步的坐标；(x′_k,y′_k)为惯性运动轨迹的相应时刻步的坐标；(x′_i+n+1,y′_i+n+1)为t_i+n+1时刻惯性运动轨迹的步的坐标。Among them, (x″′_k , y″′_k ) is the position of the post-correction step of the inertial motion trajectory; (x″_i , y″_i ) is the position coordinate point of the visual motion trajectory at time t_i , (x″_{i+n +1} , y″_i+n+1 ) is the position coordinate point of the visual motion trajectory at time t_i+n+1 , (x′_i+1 , y′_i+1 ) is the coordinate point of the inertial motion trajectory at time t_i+1 Step coordinates; (x′_k , y′_k ) are the coordinates of the step at the corresponding moment of the inertial motion trajectory; (x′_i+n+1 , y′_i+n+1 ) is the inertia moment at t_i+n+1 The coordinates of the steps of the motion trajectory.

6.根据权利要求1或2或3所述的幼儿区域活动监测方法，其特征在于，在所述步骤D中，根据特征值评估幼儿的性格喜好从而生成每个幼儿的性格喜好报告的操作包括：6. The method for monitoring regional activity of young children according to claim 1, 2 or 3, characterized in that, in the step D, the operation of evaluating the child's character preference according to the characteristic value so as to generate each child's character preference report comprises the following steps: :

根据特征值评估幼儿的专注度；Evaluate young children's concentration according to eigenvalues;

根据特征值评估幼儿的活跃度；Evaluate the activity of young children according to eigenvalues;

根据特征值评估幼儿对各子空间的偏好度；Evaluate children's preference for each subspace according to eigenvalues;

根据特征值评估各子空间的吸引力；Evaluate the attractiveness of each subspace according to the eigenvalues;

将幼儿的专注度、活跃度、各子空间的偏好度和各子空间的吸引力汇总成性格喜好报表。The children's concentration, activity, preference of each subspace and attractiveness of each subspace are aggregated into a personality preference report.

7.根据权利要求6所述的幼儿区域活动监测方法，其特征在于，在所述步骤D中，根据特征值评估幼儿的专注度的操作包括：7. The method for monitoring the activity of a child's area according to claim 6, wherein, in the step D, the operation of evaluating the child's concentration according to the characteristic value comprises:

根据各子空间出入口的坐标点信息，从视觉-惯性运动轨迹中提取幼儿进入或离开各子空间出入口的时间值；According to the coordinate point information of the entrance and exit of each subspace, extract the time value of the child entering or leaving the entrance and exit of each subspace from the visual-inertial motion trajectory;

将进入和离开子空间出入口的时间值进行相减计算得到幼儿在各子空间内的停留时长T_Kid；The time value of entering and leaving the entrance and exit of the subspace is calculated by subtraction to obtain the stay duration T_Kid of the child in each subspace;

将停留时长T_Kid记录到以幼儿姓名、子空间名称、日期、进入子空间的时间值和停留时长建立的表格中；Record the stay time T_Kid in the table created with the child's name, subspace name, date, time value of entering the subspace and stay time;

将表格中同一幼儿的所有停留时长T_Kid进行比较，将其中最大的值记为幼儿最大单次停留时长T_KidMax，进而根据停留时长T_KidMax来确定幼儿的专注度。Compare all the stay time T_Kid of the same child in the table, and record the largest value as the child's maximum single stay time T_KidMax , and then determine the child's concentration according to the stay time T_KidMax .

8.根据权利要求6所述的幼儿区域活动监测方法，其特征在于，在所述步骤D中，根据特征值评估幼儿的活跃度的操作包括：8. The method for monitoring the activity of a child's area according to claim 6, wherein, in the step D, the operation of evaluating the activity of the child according to the characteristic value comprises:

将由惯性导航设备中的加速度计采集记录的由时间参数和加速度值组成的加速度数组根据进入和离开某个子空间的时间值进行切分，提取幼儿在某个子空间的加速度数组，将某个子空间的加速度数组中的加速度值进行平均化处理，得到幼儿单区域加速度平均值；The acceleration array composed of time parameters and acceleration values collected and recorded by the accelerometer in the inertial navigation device is divided according to the time value of entering and leaving a certain subspace, and the acceleration array of the child in a certain subspace is extracted. The acceleration values in the acceleration array are averaged to obtain the average value of acceleration in a single area of the child;

将不同子空间的幼儿单区域加速度平均值进行加权平均从而得到幼儿加速度平均值，进而根据幼儿加速度平均值确定幼儿的活跃度。The average value of the acceleration of the infant in different subspaces is weighted and averaged to obtain the average value of the acceleration of the infant, and then the activity of the infant is determined according to the mean value of the acceleration of the infant.

9.根据权利要求6所述的幼儿区域活动监测方法，其特征在于，在所述步骤D中，根据特征值评估幼儿对各子空间的偏好度的操作包括：9. The method for monitoring regional activity of young children according to claim 6, wherein, in the step D, the operation of evaluating the preference of each subspace by the young child according to the eigenvalues comprises:

将同一个幼儿在同一个子空间的停留时长T_Kid进行累加获得幼儿单区域累计停留时长T_KidArea；Accumulate the stay time T_Kid of the same child in the same subspace to obtain the cumulative stay time T_KidArea of the child in a single area;

将不同子空间的幼儿单区域累计停留时长T_KidArea进行比较从而确认幼儿对各子空间的偏好度。The cumulative stay time T_KidArea of children in different subspaces is compared to confirm the children's preference for each subspace.

10.根据权利要求6所述的幼儿区域活动监测方法，其特征在于，在所述步骤D中，根据特征值评估各子空间的吸引力的操作包括：10. The method for monitoring regional activity of young children according to claim 6, wherein, in the step D, the operation of evaluating the attractiveness of each subspace according to the feature value comprises:

根据各子空间出入口的坐标点信息，从视觉-惯性运动轨迹中提取各幼儿进入或离开各子空间出入口的时间值；According to the coordinate point information of each subspace entrance and exit, extract the time value of each child entering or leaving each subspace entrance and exit from the visual-inertial motion trajectory;

将进入和离开子空间出入口的时间值进行相减计算得到各幼儿在各子空间内的停留时长T_Kid；The time values entering and leaving the entrance and exit of the subspace are subtracted and calculated to obtain the stay duration T_Kid of each child in each subspace;

将所有幼儿在同一子空间的停留时长T_Kid进行相加计算获得全班单区域累计停留时长T_ClassAreaSum；The total stay time T_Kid of all children in the same subspace is calculated to obtain the cumulative stay time T_ClassAreaSum of the whole class in a single area;

将全班单区域累计停留时长T_ClassAreaSum与该子空间预设的上限人数和该子空间的活动总时长进行相除，获得该子空间的满座率；Divide the cumulative stay time T_ClassAreaSum of the whole class in a single area by the preset upper limit of the number of people in the subspace and the total activity time of the subspace to obtain the full occupancy rate of the subspace;

将全班单区域累计停留时长T_ClassAreaSum与进入该子空间的幼儿总数进行相除得到全班单区域平均停留时长T_{ClassAreaAver}；Divide the cumulative stay time T_ClassAreaSum in the single area of the whole class by the total number of children entering the subspace to obtain the average stay time T_{ClassAreaAver} in the single area of the whole class;

将子空间的满座率小于90％的子空间判定为非热门区域并按全班单区域累计停留时长T_ClassAreaSum进行排序，获得非热门区域的吸引力排序表；Determining the subspaces with a full occupancy rate of less than 90% as non-popular areas and sorting them according to the cumulative stay time T_ClassAreaSum of the whole class single area to obtain the attractiveness ranking table of non-popular areas;

将子空间的满座率大于或等于90％的子空间判定为热门区域并按全班单区域平均停留时长T_{ClassAreaAver}进行排序，获得热门区域的吸引力排序表；Determine the subspaces whose full occupancy rate is greater than or equal to 90% as popular areas and sort them according to the average stay time T_{ClassAreaAver} of the whole class to obtain the attractiveness ranking table of popular areas;

通过非热门区域的吸引力排序表和热门区域的吸引力排序表确认各子空间对幼儿的吸引力。Confirm the attractiveness of each subspace to young children through the attractiveness ranking table of non-popular areas and the attractiveness ranking table of popular areas.