Study On Automatic Target Scoring Based On Image Processing Technology

With the continuous development of science and technologies,human's life and works become more and more automated.In order to reduce the time of target scoring,and improve the efficiency and fairness of competitions and trainings,the automatic target scoring mode is gradually replacing the traditional manual target scoring mode.This paper,analyzing the present situation of researches on the automatic target scoring system at home and abroad,focuses on the study of automatic target scoring system of light weapons shooting of chest ring targets based on image processing technology.First of all,according to the process of traditional light weapons shooting training and target scoring,combined with the existing designs of the automatic scoring system based on image processing technology,developed an overall design communicating with wireless.Analyzing the features of target image acquired in the complex environment outdoors,developed the overall image processing module of the automatic target scoring system,namely: first,locate the effective target area in the target image,extract the effective target image.Then mark each region with a corresponding score according to the characteristics of the effective target image.Finally identify the bullet hole,determine the score according to the marked area of the center position of the bullet hole.Then,focusing on the image-processing algorithms of the automatic target scoring system,analyses and researches have been done: 1.Analyzing the characteristics of the target image acquired in complex environment outdoors,combined with the traditional definitions and characteristics of the H,S,I component in HSI color space of an image,defined a concept of away-from-red saturation,proposed an algorithm of target area positioning and effective target extraction based on color and edge features.2.Analyzing the color feature of the target image,proposed a region labeling algorithm based on gradual seed filling,to label each area using characteristics of the target image directly.3.Analyzing the mutation characteristics of bullet holes in gray and color,proposed an algorithm of bullet holes' recognition based on the mutation features of local pixels.By checking which area labeled the center position of each bullet hole is,determined the score.Finally,according to the image-processing algorithms proposed in this paper,using GUIDE software development environment of MATLAB,developed a software platform of the automatic target scoring system which is relatively complete,simple,feasible and easy to use,then tested the performance of the system with 30 target images acquired in complex environment outdoors different in size and imaging angle.The result shows that the system has advantages of strong adaptability to different environment,high stability,good real-time,and fewer restriction on the image acquisition hardware equipment,which can meet the requirements of realistic application of target scoring,and has certain reference value for practical applications.