Automatic Target-scoring Model based on Image Processing

At present, in the field of military shooting, most of the target-scoring methods are artificial, acoustics-based, optics-based, and so on. These kinds of methods have their own disadvantages respectively such as high cost, time-consuming and low accuracy. To solve these problems, starting from the field of computer vision, combined with the image processing algorithms, in this paper, we propose an automatic target-scoring model. Firstly, image correction and subtraction operations are performed on the two images before and after shooting. Then, image processing operations such as binarization, opening and closing operations, and Canny edge detection are performed on the images that have eliminated most of the background interference, and a clear image of a single bullet hole is obtained. Finally, the Hough transform algorithm is used to identify the bullet holes, the Euclidean distance between the center of the bullet hole and the target surface is calculated, so the final target score is obtained. The testing images are all taken at a fixed angle of view 50 meters away from the targets. The experimental results show that the algorithm can accurately identify the location of bullet holes, with fast processing and more low cost than other algorithms.