Application of Maximum Stability Extreme Value Region in Dynamic Target Detection

In the dynamic laser imaging target recognition system, the target resolution of different distance images is different, which increases the difficulty of regional feature detection, easily causes false detection or missed detection, and seriously affects the accuracy of target recognition. In view of this situation, using the Maximally Stable Extremal Regions (MSER) algorithm to detect the characteristics of the target area in different distance images, the test results are expressed by ellipse fitting and compared with the original target image test results. Experiments show that the proposed algorithm can effectively detect the regional features of the target under dynamic laser imaging and improve the accuracy of target detection.