Research On Application Of Image Gray Similarity Projection In Low-altitude Target Detection

With the gradual opening of low-altitude airspace,the types and number of low-altitude aircraft continue to increase,“black fly” incidents occur frequently,which pose a threat to social public security,and become a problem that governments and the public are worried about.Therefore,there is an urgent need to study the reliable detection technology for low-altitude targets to provide a strong guarantee for social public security.Aiming at the existing problems of low-altitude target detection,an image gray similarity projection detection method for low-altitude target is proposed in this thesis,and the computer vision detection of low-altitude target is realized by using the upper computer human-computer interaction software system.Firstly,the related theory of video image processing is introduced in this thesis,including image graying,image filtering and image morphological processing.Then the basic principles of three typical traditional moving target detection methods: background difference method,frame difference method and optical flow method is described,and the defects of background difference method and frame difference method in detecting low-altitude targets is analyzed.Image gray similarity projection target detection method is based on the different characteristics of the ground background and sky region in the row gray similarity projection curve to achieve the segmentation of the sky area,so as to remove the influence of the complex ground background on the target detection in sky area;according to the difference in gray value between the target,cloud and sky background in sky area,the number of objects is determined by the number of extreme points of row and column projection curves of pixel gray similarity,and the image is segmented by the center of adjacent extreme points to achieve the pixel coordinate matching of moving objects.On the basis of the gray similarity projection method,the Euclidean distance comparison between the pixel coordinates of the target in the adjacent frames is used to achieve the trajectory matching of multiple targets.Then the video reading and display,as well as the detection,positioning and trajectory drawing of moving objects in the video is completed through the development of the upper computer human-computer interaction software.Finally,the feasibility of image gray similarity projection method is verified by experiment and the upper computer system is tested.The performance of gray similarityprojection method and traditional method is compared and analyzed.The experimental results show that the image gray similarity projection method can effectively detect low-altitude targets with complex ground background and environmental light changes.Compared with the traditional method,the detection speed is increased by 14%,the recognition rate is 97.4%,and the false detection rate is only 2.3%,which provides an effective technical means for the follow-up application research.