Automatic Target-scoring System Based On Video Image Analysis

With the rapid development of modern science and technology, the degree of automation becomes more and more advanced in all walks. Shooting target-scoring technology has also evolved from artificial into the automatic scoring stage. Compared with the traditional manual scoring method, automatic scoring methods can effectively overcome the artificial human error, fit the needs of military modernization, and raise the level of army shooting training. Therefore, research on automatic target-scoring system which has good performance has important value in engineering.In view of existing automatic target-score system in shooting accuracy and lack of adaptability to the environment, this thesis has developed a set of automatic target-score system based on video analysis. The system is based on the TMS320DM8127 processor, which implements the video encoding function on M3 core, implements the automatic target-score business on A8 core and implements the automatic target-score function on DSP core. The system functions include vibration testing, target capture and live business of shooting. Based on multilayer framework concept, the software can be dived into 4 layers, including hardware layer, driven layer, interface layer and application layer. The Algorithm of the auto target-score can be divided into 3 stages, including intelligent classifier training, target identification and target-scoring, using dependent layer, add layer and application layer of 3-tier architecture to achieve.The results show that the target-scoring system in this thesis has high accuracy and can adapts the complex environmental conditions, and it can satisfy the live-fire requirements.