技术领域Technical field
本申请涉及涉及图像处理技术领域，尤其涉及一种红外故障检测识别算法。The present application relates to the field of image processing technology, and in particular to an infrared fault detection and identification algorithm.
背景技术Background technique
物体表面温度如果超过绝对零度即会辐射出电磁波，红外线和人们熟知的可见光都属于电磁波，随着温度变化，电磁波的辐射强度与波长分布特性也随之改变，波长介于0.75μm到1000μm间的电磁波称为“红外线”，红外线在地表传送时，会受到大气组成物质(特别是H2O、CO2、CH4、N2O、O3等)的吸收，强度明显下降，仅在中波3μ～5μm及长波8～12μm的两个波段有较好的穿透率(Transmission)，大部份的红外热像仪就是针对这两个波段进行检测。电气设备红外图像热故障诊断研究现状红外成像技术始20世纪30年代，随后较长的一段时间里，红外成像技术主要应用在军事领域中，使其得到了快速的发展，而后被逐渐应用到国民经济的各个领域中去，如工业监测、安防保卫、医学诊断等领域。70年代，瑞典国家电力局使用集成红外成像系统的汽车和直升机对变电站电气设备进行巡检。1990年的国际大电网会议对红外诊断技术给予了足够的重视和充分的肯定。2006年FLIR推出ThermaCAMP640是全球首台选用焦平面探测器的热像仪，并将其应用到电气设备巡检任务中。在国内，电气设备红外检测技术始于70年代初，并在军用和民用两方面都取得了良好的效果。随着信息处理和其他相关技术的成熟，红外诊断技术正向数字化、智能化的方向发展，并且已经结合数字图像处理技术应用于电气设备的热故障诊断。If the surface temperature of an object exceeds absolute zero, it will radiate electromagnetic waves. Infrared rays and the well-known visible light are both electromagnetic waves. As the temperature changes, the radiation intensity and wavelength distribution characteristics of the electromagnetic waves also change. The wavelength is between 0.75μm and 1000μm. Electromagnetic waves are called "infrared rays". When infrared rays are transmitted on the earth's surface, they will be absorbed by atmospheric components (especially H2O, CO2, CH4, N2O, O3, etc.), and the intensity will drop significantly. The two 12μm bands have better transmission, and most infrared thermal imaging cameras detect these two bands. Current status of research on thermal fault diagnosis of infrared images of electrical equipment. Infrared imaging technology began in the 1930s. For a long period of time, infrared imaging technology was mainly used in the military field, causing rapid development and then gradually being applied to national applications. Go to various fields of the economy, such as industrial monitoring, security, medical diagnosis and other fields. In the 1970s, the Swedish National Electricity Board used cars and helicopters with integrated infrared imaging systems to inspect electrical equipment in substations. The 1990 International Conference on Large Power Grids paid sufficient attention and full recognition to infrared diagnostic technology. In 2006, FLIR launched ThermaCAMP640, the world's first thermal imaging camera using a focal plane detector, and applied it to electrical equipment inspection tasks. In China, infrared detection technology for electrical equipment began in the early 1970s and has achieved good results in both military and civilian applications. With the maturity of information processing and other related technologies, infrared diagnostic technology is developing in the direction of digitalization and intelligence, and has been combined with digital image processing technology for thermal fault diagnosis of electrical equipment.
特征选择是指去掉无关特征，保留相关特征的过程，也可以认为是从所有的特征中选择一个最好的特征子集。特征选择本质上可以认为是降维的过程。在计算机视觉中，主要提取图像特征。图像特征主要有图像的颜色特征、纹理特征、形状特征和空间关系特征。颜色特征是一种全局特征，描述了图像或图像区域所对应的景物的表面性质。纹理特征也是一种全局特征，它也描述了图像或图像区域所对应景物的表面性质。但由于纹理只是一种物体表面的特性，并不能完全反映出物体的本质属性，所以仅仅利用纹理特征是无法获得高层次图像内容的。与颜色特征不同，纹理特征不是基于像素点的特征，它需要在包含多个像素点的区域中进行统计计算。形状特征有两类表示方法，一类是轮廓特征，另一类是区域特征。图像的轮廓特征主要针对物体的外边界，而图像的区域特征则关系到整个形状区域所谓空间关系，是指图像中分割出来的多个目标之间的相互的空间位置或相对方向关系，这些关系也可分为连接/邻接关系、交叠/重叠关系和包含/包容关系等。通常空间位置信息可以分为两类：相对空间位置信息和绝对空间位置信息。前一种关系强调的是目标之间的相对情况，如上下左右关系等，后一种关系强调的是目标之间的距离大小以及方位。常见的图像特征提取方式有：SIFT、HOG、SURF、ORB、LBP、HAAR和卷积神经网络等，所以一种能在复杂背景下对红外图像精确的计算的方法尤为重要。Feature selection refers to the process of removing irrelevant features and retaining relevant features. It can also be considered as selecting the best feature subset from all features. Feature selection can essentially be considered as a process of dimensionality reduction. In computer vision, image features are mainly extracted. Image features mainly include color features, texture features, shape features and spatial relationship features of the image. Color feature is a global feature that describes the surface properties of the scene corresponding to the image or image area. Texture features are also global features, which also describe the surface properties of the scene corresponding to the image or image area. However, since texture is only a characteristic of the surface of an object and cannot fully reflect the essential properties of the object, it is impossible to obtain high-level image content simply by using texture features. Unlike color features, texture features are not pixel-based features and require statistical calculations in an area containing multiple pixels. There are two types of representation methods for shape features, one is contour features and the other is regional features. The contour features of the image are mainly aimed at the outer boundary of the object, while the regional features of the image are related to the entire shape area. The so-called spatial relationship refers to the mutual spatial position or relative direction relationship between multiple targets segmented in the image. These relationships It can also be divided into connection/adjacency relationships, overlapping/overlapping relationships, inclusion/containment relationships, etc. Generally, spatial location information can be divided into two categories: relative spatial location information and absolute spatial location information. The former relationship emphasizes the relative situation between targets, such as the up, down, left, and right relationships, etc., while the latter relationship emphasizes the distance and orientation between targets. Common image feature extraction methods include: SIFT, HOG, SURF, ORB, LBP, HAAR and convolutional neural network, etc. Therefore, a method that can accurately calculate infrared images in complex backgrounds is particularly important.
发明内容Contents of the invention
有鉴于此，本发明的目的是提供了一种红外故障检测识别算法，通过计算方差加权信息熵衡量红外图像的复杂度，然后分割提取故障区域，引入规则和分隔与膨胀的迭代确定故障位置和尺寸信息，通过此方法提供了一种有效降低误检率和提高精确率的复杂背景下的红外图像处理的方法，增强了算法的自适应性,频域检测时域和识别的结合，提高了运算效率，降低了误检率，提高了准确率，增强了其自适应性In view of this, the purpose of the present invention is to provide an infrared fault detection and identification algorithm, which measures the complexity of the infrared image by calculating the variance weighted information entropy, then segments and extracts the fault area, and introduces rules and iterations of separation and expansion to determine the fault location and Size information, this method provides a method of infrared image processing under complex backgrounds that effectively reduces the false detection rate and improves accuracy, enhances the adaptability of the algorithm, and combines frequency domain detection with time domain and recognition, improving Operation efficiency reduces false detection rate, improves accuracy and enhances its adaptability.
本发明通过以下技术手段解决上述技术问题：The present invention solves the above technical problems through the following technical means:
本发明提供了一种红外故障检测识别算法，具体按以下步骤执行：The present invention provides an infrared fault detection and identification algorithm, which is specifically executed according to the following steps:
S1：计算方差加权信息熵衡量红外图像的复杂度，并根据复杂度变量建立与截止频率的映射关系；S1: Calculate the variance weighted information entropy to measure the complexity of the infrared image, and establish a mapping relationship with the cutoff frequency based on the complexity variable;
S2：在红外图像的频域上，采用自适应Butterworth滤波技术来抑制背景，分割并提取故障潜在区域，并滤除指定比例的像素；S2: In the frequency domain of the infrared image, adaptive Butterworth filtering technology is used to suppress the background, segment and extract potential fault areas, and filter out a specified proportion of pixels;
S3：根据图像分割处理后的像素分布特点，利用直方图计算特征数据，引入规则判断图像是否存在故障；S3: Based on the pixel distribution characteristics after image segmentation, use histograms to calculate feature data, and introduce rules to determine whether there is a fault in the image;
S4：在红外图像的时域上，分块滤除指定比例的像素，消除干扰与故障的连通性；S5：通过分割与膨胀的迭代进一步确定故障的信息，识别并获取故障位置和尺寸信息。S4: In the time domain of the infrared image, filter out a specified proportion of pixels in blocks to eliminate interference and connectivity of faults; S5: Further determine the fault information through iterations of segmentation and expansion, and identify and obtain fault location and size information.
进一步，在步骤S1中，在步骤S1中，计算方差加权信息熵衡量图像复杂度，并建立与截止频率的映射关系；使用在红外图像的频域上做Butterworth自适应高通滤波来抑制大范围背景。Further, in step S1, in step S1, the variance weighted information entropy is calculated to measure the image complexity, and a mapping relationship with the cutoff frequency is established; Butterworth adaptive high-pass filtering is used in the frequency domain of the infrared image to suppress the large-scale background .
进一步，在步骤S2中，对于S1的处理结果，借助直方图计算可区分的数据判定图像中是否存在故障；取结果图像的灰度最大值p_max，非零像素平均值nz_p_aver，非零像素最小值nz_p_min，将这三个数据组合如下：Further, in step S2, for the processing result of S1, use the histogram to calculate distinguishable data to determine whether there is a fault in the image; take the maximum gray value p_max of the result image, the average non-zero pixel value nz_p_aver, and the minimum non-zero pixel value nz_p_min, combine these three data as follows:
(1)p_max；(2)p_max-nz_p_aver；(3)p_max-nz_p_min；(1)p_max; (2)p_max-nz_p_aver; (3)p_max-nz_p_min;
结合阈值设计判断条件，可具有明显区分性。Combined with the threshold design judgment conditions, it can be clearly differentiated.
进一步，在步骤S3中，当故障与背景高亮区域重叠时，二者之间差异不容易区分；通过分块处理，可以滤除更多的高亮非故障区域，降低误检率。Further, in step S3, when the fault overlaps with the background highlighted area, the difference between the two is not easy to distinguish; through block processing, more highlighted non-fault areas can be filtered out and the false detection rate can be reduced.
进一步，在步骤S4中，对于S3的结果图像，通过更新阈值做二值化和膨胀处理迭代，可使得最终的分割结果只有一个目标，可用轮廓个数来代替目标个数。Further, in step S4, for the result image of S3, the binarization and expansion processing iterations are performed by updating the threshold, so that the final segmentation result has only one target, and the number of contours can be used instead of the number of targets.
进一步，在步骤S5中膨胀的迭代具体按以下步骤执行：Further, the iteration of expansion in step S5 is specifically performed as follows:
S1：计算当前图像非零像素灰度平均值nz_aver作为阈值thresh；S1: Calculate the non-zero pixel gray average value nz_aver of the current image as the threshold thresh;
S2：以阈值thresh分割图像f，更新阈值thresh加1；S2: Segment the image f with the threshold thresh, and update the threshold thresh plus 1;
S3：图像f膨胀运算；S3: Image f expansion operation;
S4:检测图像轮廓并计算个数con_number；S4: Detect image contours and calculate the number con_number;
S5:判断con_number＝1？,是则终止计算，返回阈值thresh，否则继续判断thresh＝255？，是则返回special_thresh，否则返回步骤S2；S5: Determine con_number=1? , if yes, terminate the calculation and return the threshold thresh, otherwise continue to judge threshold=255? , if yes, return special_thresh, otherwise return to step S2;
S6:结束。S6: End.
本发明的一种红外故障检测识别算法，通过计算方差加权信息熵衡量红外图像的复杂度，然后分割提取故障区域，引入规则和分隔与膨胀的迭代确定故障位置和尺寸信息，通过此方法提供了一种有效降低误检率和提高精确率的红外图像处理的方法，增强了算法的自适应性,频域检测时域和识别的结合，提高了运算效率，降低了误检率，该算法可以在大幅度降低误检率的情况下保证正确率几乎不变，提高了准确率，增强了其自适应性。这使得红外故障识别进一步的贴切其使用场景，加快无人机巡检电力设备的落地应用。An infrared fault detection and identification algorithm of the present invention measures the complexity of the infrared image by calculating the variance weighted information entropy, and then segments and extracts the fault area. It introduces rules and iteratively determines the fault location and size information through separation and expansion. This method provides An infrared image processing method that effectively reduces the false detection rate and improves accuracy. It enhances the adaptability of the algorithm. The combination of frequency domain detection, time domain and recognition improves computing efficiency and reduces the false detection rate. This algorithm can While greatly reducing the false detection rate, the accuracy rate remains almost unchanged, the accuracy rate is improved, and its adaptability is enhanced. This makes infrared fault identification more suitable for its use scenarios and accelerates the application of drones for inspecting power equipment.
附图说明Description of the drawings
图1为原始红外图像；Figure 1 is the original infrared image;
图2为原始红外灰度直方图；Figure 2 is the original infrared grayscale histogram;
图3为自适应Butterworth滤波图；Figure 3 shows the adaptive Butterworth filter diagram;
图4为频域滤波结果图的灰度直方图；Figure 4 shows the grayscale histogram of the frequency domain filtering result image;
图5为指定比例的阈值分割结果图；Figure 5 shows the threshold segmentation result of a specified ratio;
图6为时域分割图；Figure 6 is a time domain segmentation diagram;
图7为迭代流程图；Figure 7 is an iterative flow chart;
图8为故障识别图；Figure 8 is a fault identification diagram;
其中：图八的矩形框内的区域为通过本算法识别为的故障区域。Among them: the area within the rectangular box in Figure 8 is the fault area identified by this algorithm.
具体实施方式Detailed ways
以下将结合附图和具体实施例对本发明进行详细说明，显然，所描述的实施例仅仅只是本申请一部分实施例，而不是全部的实施例，基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, those of ordinary skill in the art All other embodiments obtained without creative efforts fall within the scope of protection of this application.
如图2所示，本发明的一种红外故障检测识别算法，本发明公开了一种复杂背景下红外故障检测识别算法，该算法包括以下步骤：As shown in Figure 2, an infrared fault detection and identification algorithm of the present invention is disclosed. The invention discloses an infrared fault detection and identification algorithm under a complex background. The algorithm includes the following steps:
S1：计算方差加权信息熵衡量红外图像的复杂度，并根据复杂度变量建立与截止频率的映射关系；S1: Calculate the variance weighted information entropy to measure the complexity of the infrared image, and establish a mapping relationship with the cutoff frequency based on the complexity variable;
S2：在红外图像的频域上，采用自适应Butterworth滤波技术来抑制背景，分割并提取故障潜在区域，并滤除指定比例的像素；S2: In the frequency domain of the infrared image, adaptive Butterworth filtering technology is used to suppress the background, segment and extract potential fault areas, and filter out a specified proportion of pixels;
S3：根据图像分割处理后的像素分布特点，利用直方图计算特征数据，引入规则判断图像是否存在故障；S3: Based on the pixel distribution characteristics after image segmentation, use histograms to calculate feature data, and introduce rules to determine whether there is a fault in the image;
S4：在红外图像的时域上，分块滤除指定比例的像素，消除干扰与故障的连通性；S5：通过分割与膨胀的迭代进一步确定故障的信息，识别并获取故障位置和尺寸信息。S4: In the time domain of the infrared image, filter out a specified proportion of pixels in blocks to eliminate interference and connectivity of faults; S5: Further determine the fault information through iterations of segmentation and expansion, and identify and obtain fault location and size information.
在实施例中，如步骤S1所述，基于自适应Butterworth高通滤波的红外小目标检测方法。一个具有良好滤波性能的2阶频域Butterworth高通滤波器,其传递函数可写为：In the embodiment, as described in step S1, an infrared small target detection method is based on adaptive Butterworth high-pass filtering. A second-order frequency domain Butterworth high-pass filter with good filtering performance, its transfer function can be written as:
式中D0为截止频率，D(u，v)为频谱中的点，(u，v)为到频谱中心的欧式距离，计算公式为：In the formula, D0 is the cut-off frequency, D(u, v) is the point in the spectrum, (u, v) is the Euclidean distance to the center of the spectrum, and the calculation formula is:
式中(P，Q)表示频谱中心点，用像素灰度值分别求取4块的方差加权信息熵，并计算最大值作为图像复杂度的结果，公式如下：In the formula (P, Q) represents the center point of the spectrum, the variance weighted information entropy of the four blocks is calculated using the pixel gray value, and the maximum value is calculated as the result of the image complexity. The formula is as follows:
H(s)max＝max(H(s)tl,H(s)tr,H(s)bl,H(s)br)...式7H(s)max＝max(H(s)tl ,H(s)tr ,H(s)bl ,H(s)br )...Equation 7
式中H(s)tl为图像左上角方差加权信息熵，H(s)tr为图像右上角方差加权信息熵，H(s)bl为图像左下角方差加权信息熵，H(s)br为图像右下角方差加权信息熵，H(s)max为代表图像复杂度的熵值，为每一块图像对应的灰度值平均值。In the formula, H(s)tl is the variance-weighted information entropy of the upper left corner of the image, H(s)tr is the variance-weighted information entropy of the upper right corner of the image, H(s)bl is the variance-weighted information entropy of the lower left corner of the image, H(s)br is Variance weighted information entropy in the lower right corner of the image, H(s)max is the entropy value representing the complexity of the image, is the average gray value corresponding to each image.
本实施例中，如步骤S2所述，分析滤波分割后的图像的灰度最大值p_max，非零像素平均值nz_p_aver，非零像素最小值nz_p_min，In this embodiment, as described in step S2, the maximum gray value p_max, the average non-zero pixel value nz_p_aver, and the minimum non-zero pixel value nz_p_min of the filtered and segmented image are analyzed.
将这三个数据组合计算如下：The combination of these three data is calculated as follows:
p_maxp_max
p_max-nz_p_averp_max-nz_p_aver
p_max-mz_p_minp_max-mz_p_min
每幅图像的三个参数为一组数据。通过组合参数来生成判断规则，每一条规则由三个条件组成，其形式为The three parameters of each image are a set of data. Judgment rules are generated by combining parameters. Each rule consists of three conditions, in the form of
fault_flag＝flag1&&flag2&&flag3...式9fault_flag＝flag1 &&flag2 &&flag3 ...Equation 9
式中t1，t2，t3为每一个参数对应的阈值，flag1，flag2，flag3为每个条件的执行结果，fault_flag为最终的判断结果，为true则代表图像存在故障，为false不存在故障。In the formula, t1 , t2 , and t3 are the thresholds corresponding to each parameter, flag1 , flag2 , and flag3 are the execution results of each condition. fault_flag is the final judgment result. If it is true, it means that there is a fault in the image, as false There is no fault.
本实施例中，如步骤S3所述，选取合适的图像分块参数K。K的选取一般由实验或经验获得。如使用K＝4的实验数据，如图6。如步骤4所示，对步骤S3的结果图像迭代计算故障信息。In this embodiment, as described in step S3, an appropriate image block parameter K is selected. The selection of K is generally obtained by experiment or experience. For example, use the experimental data of K=4, as shown in Figure 6. As shown in step 4, the fault information is iteratively calculated for the result image of step S3.
以上实施例仅用以说明本发明的技术方案而非限制，尽管参照较佳实施例对本发明进行了详细说明，本领域的普通技术人员应当理解，可以对本发明的技术方案进行修改或者等同替换，而不脱离本发明技术方案的宗旨和范围，其均应涵盖在本发明的权利要求范围当中。本发明未详细描述的技术、形状、构造部分均为公知技术。The above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently substituted. Without departing from the purpose and scope of the technical solutions of the present invention, they should all be covered by the claims of the present invention. The technology, shape, and structural parts not described in detail in the present invention are all known technologies.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910930055.5ACN110660065B (en) | 2019-09-29 | 2019-09-29 | Infrared fault detection and identification algorithm |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910930055.5ACN110660065B (en) | 2019-09-29 | 2019-09-29 | Infrared fault detection and identification algorithm |
| Publication Number | Publication Date |
|---|---|
| CN110660065A CN110660065A (en) | 2020-01-07 |
| CN110660065Btrue CN110660065B (en) | 2023-10-20 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910930055.5AActiveCN110660065B (en) | 2019-09-29 | 2019-09-29 | Infrared fault detection and identification algorithm |
| Country | Link |
|---|---|
| CN (1) | CN110660065B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111260624B (en)* | 2020-01-15 | 2023-05-26 | 陕西师范大学 | A Multi-Scale Infrared Small Target Detection Method Based on Feature Matching and Scale Selection Strategy |
| CN111768376A (en)* | 2020-06-24 | 2020-10-13 | 山东科技大学 | An eddy current thermal imaging edge detection method, system, storage medium and application |
| CN113139576B (en)* | 2021-03-22 | 2024-03-12 | 广东省科学院智能制造研究所 | A deep learning image classification method and system combining image complexity |
| CN114548446B (en)* | 2022-04-25 | 2022-08-02 | 国网山东省电力公司潍坊市寒亭区供电公司 | Power equipment detection system and method based on artificial intelligence |
| CN117911401B (en)* | 2024-03-15 | 2024-06-21 | 国网山东省电力公司泗水县供电公司 | A method, system, storage medium and device for detecting faults of electric power equipment |
| CN118310635B (en)* | 2024-06-11 | 2024-08-13 | 山西创芯光电科技有限公司 | Focal plane detection method and system for infrared detector |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005069197A1 (en)* | 2003-12-31 | 2005-07-28 | Lockheed Martin Missiles & Fire Control | A method and system for adaptive target detection |
| JP2010166143A (en)* | 2009-01-13 | 2010-07-29 | Ricoh Co Ltd | Image processing apparatus, image processing method, program, and recording medium |
| CN102034239A (en)* | 2010-12-07 | 2011-04-27 | 北京理工大学 | Local gray abrupt change-based infrared small target detection method |
| CN103198326A (en)* | 2013-04-15 | 2013-07-10 | 杭州捷尚智能电网技术有限公司 | Electric transmission line helicopter routing inspection image classification method |
| CN104657945A (en)* | 2015-01-29 | 2015-05-27 | 南昌航空大学 | Infrared small target detection method for multi-scale spatio-temporal union filtering under complex background |
| CN107784661A (en)* | 2017-09-08 | 2018-03-09 | 上海电力学院 | Substation equipment infrared image classifying identification method based on region-growing method |
| CN108154087A (en)* | 2017-12-08 | 2018-06-12 | 北京航天计量测试技术研究所 | A kind of matched infrared human body target detection tracking method of feature based |
| CN108665443A (en)* | 2018-04-11 | 2018-10-16 | 中国石油大学(北京) | A kind of the infrared image sensitizing range extracting method and device of mechanical equipment fault |
| CN109712158A (en)* | 2018-11-23 | 2019-05-03 | 山东航天电子技术研究所 | A kind of infrared small target catching method based on target background pixel statistical restraint |
| CN109993736A (en)* | 2019-03-29 | 2019-07-09 | 浙江大学 | A thermal imaging segmentation method for electrical equipment based on improved weighted otsu algorithm |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105678310B (en)* | 2016-02-03 | 2019-08-06 | 北京京东方多媒体科技有限公司 | Infrared thermal image contour extraction method and device |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005069197A1 (en)* | 2003-12-31 | 2005-07-28 | Lockheed Martin Missiles & Fire Control | A method and system for adaptive target detection |
| JP2010166143A (en)* | 2009-01-13 | 2010-07-29 | Ricoh Co Ltd | Image processing apparatus, image processing method, program, and recording medium |
| CN102034239A (en)* | 2010-12-07 | 2011-04-27 | 北京理工大学 | Local gray abrupt change-based infrared small target detection method |
| CN103198326A (en)* | 2013-04-15 | 2013-07-10 | 杭州捷尚智能电网技术有限公司 | Electric transmission line helicopter routing inspection image classification method |
| CN104657945A (en)* | 2015-01-29 | 2015-05-27 | 南昌航空大学 | Infrared small target detection method for multi-scale spatio-temporal union filtering under complex background |
| CN107784661A (en)* | 2017-09-08 | 2018-03-09 | 上海电力学院 | Substation equipment infrared image classifying identification method based on region-growing method |
| CN108154087A (en)* | 2017-12-08 | 2018-06-12 | 北京航天计量测试技术研究所 | A kind of matched infrared human body target detection tracking method of feature based |
| CN108665443A (en)* | 2018-04-11 | 2018-10-16 | 中国石油大学(北京) | A kind of the infrared image sensitizing range extracting method and device of mechanical equipment fault |
| CN109712158A (en)* | 2018-11-23 | 2019-05-03 | 山东航天电子技术研究所 | A kind of infrared small target catching method based on target background pixel statistical restraint |
| CN109993736A (en)* | 2019-03-29 | 2019-07-09 | 浙江大学 | A thermal imaging segmentation method for electrical equipment based on improved weighted otsu algorithm |
| Title |
|---|
| 基于复杂融合特征与灰度-纹理直方图描述子的红外弱小目标检测追踪算法;闻凯;《科学技术与工程》;20161208(第34期);88-96* |
| 基于背景复杂程度估计的红外图像预处理;杨磊等;《系统工程与电子技术》;20060820(第08期);40-43* |
| 复杂背景条件下的红外小目标检测与跟踪算法研究;杨磊;《中国优秀博硕士学位论文全文数据库（博士） 信息科技辑》;20070215;1-141* |
| Publication number | Publication date |
|---|---|
| CN110660065A (en) | 2020-01-07 |
| Publication | Publication Date | Title |
|---|---|---|
| CN110660065B (en) | Infrared fault detection and identification algorithm | |
| CN110415208B (en) | An adaptive target detection method and its device, equipment and storage medium | |
| CN111260616A (en) | Insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization | |
| CN109460764B (en) | Satellite video ship monitoring method combining brightness characteristics and improved interframe difference method | |
| CN111079518B (en) | Ground-falling abnormal behavior identification method based on law enforcement and case handling area scene | |
| CN102496016A (en) | Infrared target detection method based on space-time cooperation framework | |
| CN104504686A (en) | Hyper-spectral image abnormity detection method adopting local self-adaptive threshold segmentation | |
| CN104715474B (en) | Linear Building Detection Method for High Resolution Synthetic Aperture Radar Image Based on Marker Watershed Algorithm | |
| Li et al. | IDP-YOLOV9: Improvement of Object Detection Model in Severe Weather Scenarios from Drone Perspective. | |
| CN113962900B (en) | Infrared weak target detection method, device, equipment and medium under complex background | |
| CN107767404A (en) | A kind of remote sensing images sequence moving target detection method based on improvement ViBe background models | |
| CN104933728A (en) | Mixed motion target detection method | |
| CN112070717A (en) | Power transmission line icing thickness detection method based on image processing | |
| CN106529472B (en) | Object detection method and device based on large scale high-resolution high spectrum image | |
| CN120047440B (en) | Method, device, equipment and medium for detecting surface cracks of aviation structural component | |
| CN115311678A (en) | An Infrared Video Airport Bird Detection Method Combining Background Suppression and DCNN | |
| CN115731255A (en) | Image edge enhancement method | |
| CN108171661B (en) | Infrared target detection method based on improved Tri edge operator | |
| Zhu et al. | Shadow removal with background difference method based on shadow position and edges attributes | |
| CN112967305B (en) | Image cloud background detection method under complex sky scene | |
| CN119206530B (en) | Dynamic target identification method, device, equipment and medium for remote sensing image | |
| CN118887200B (en) | Defect identification method, defect identification device, computer device and readable storage medium | |
| CN114820384A (en) | Infrared weak and small target detection method based on local contrast and gradient | |
| CN112288780B (en) | Multi-feature dynamically weighted target tracking algorithm | |
| Wang et al. | Shadow detection and reconstruction of high-resolution remote sensing images in mountainous and hilly environments |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |