Research on UAV Small Target Segmentation Algorithm Under Complex Sky Background
In order to solve the problem of detection of small targets in the sky background,this paper first an-alyzes the characteristics of UAV small targets and interference on the basis of images captured by photoelectric de-tection equipment.Then image filtering and improved Sobel operator algorithm are used to suppress the interference of clouds and background interference.Meanwhile,morphological corrosion and swelling are introduced to enhance the signal of small targets.Finally,an adaptive thresholding target extraction method is proposed to obtain the size and position information of small targets from gray image.The experimental result shows that background suppres-sion and detection methods have strong ability to suppress background noise.In the case of RGB camera imaging,infrared imaging,high-power laser supplementing light and other various environments,target extraction method has a relatively stable effect on the location extraction of diminutive targets.
万方数据
维普
