Design of automatic target-scoring system of shooting game based on computer vision

This paper describes a machine vision system with back lighting illumination and friendly man-machine interface. Subtraction is used to segment target holes quickly and accurately. The oval obtained after tracing boundary is processed by Generalized Hough Transform to acquire the target´s center. Marked-hole´s area, perimeter and moment invariants are extracted as cluster features. The auto-scoring software, programmed by Visual C++, has successfully solved the recognition of off-target and overlapped holes through alarming surveillance and bullet tacking programs. The experimental results show that, when the target is distorted obviously, the system can recognize the overlapped holes on real time and also clusters random shape holes on the target correctly. The high accuracy, fast computing speed, easy debugging and low cost make the system can be widely used.