A bullet hole recognition method based on visual detection

Shooting Intelligent Detection System has important applications in the military industry. Existing target shooting systems mainly face problems such as high cost, complex operation, and poor environmental adaptability. Considering the above limitations, in this paper we propose a fast, low-cost, and easy-to-deploy bullet hole recognition algorithm based on visual detection, which mainly consists of four parts: image preprocessing, extraction of target center, bullet hole extraction and ring value determination. In the process of bullet hole extraction, a method for segmenting effective target surface is proposed to cope with the problem of off-target bullet holes. Our method can not only accurately localize the bullet holes, but also obtain the exact ring value at that location. In addition, due to the robust adaptability of our method, it can even be applied to outdoor scenarios.