自动报靶系统研究综述

Summary on Automatic Target-scoring System

介绍了一种基于机器视觉的自动报靶系统,该系统利用CCD摄像机不间断地对靶面进行观测,根据采集的靶面图像的特点和变化,利用计算机图像识别技术检测靶面图像中的真实弹点,然后通过弹点在靶面图中的位置来对弹点区域进行编码。在观测过程中,采用图像灰度配准与增强技术、图像的几何配准技术和小波变换等算法来解决由于光照变化、大风等干扰因素对识别精度的影响。在实际应用中,能够达到较高精度,并能大大减少错报和漏报的情况,该系统精度高、成木低、完全可靠,能够很好地应用于部队训练和射击比赛中有着重要的实际意义和良好的应用前景。

An automatic target-scoring system is proposed, which based on machine vision.CCD camera was used to observe targets' surface continuously.According to the feature and change of the images of targets' surface observed by CCD, the technology of computer pattern recognition was used to detect the position of real shell in the target surface.Then the position was described in digital by coding target surface.In the process of observation, the technology on enhancement and matching of gray-scale, image matching technology in geometry, and wavelet transform algorithms were used to solve the effect on the accuracy of the identification due to changes in light, wind, and other factors.In practice, the system can achieve ideal accuracy, reduce the misstatements and omissions greatly.It was also high-precision, low-cost,and reliable system, can be used in military drill and shooting competitions.It will be meaningful and have good prospect.