户外射击训练中采用视觉检测技术进行自动报靶, 既可防止发生训练安全事故, 还可以获得训练者在射击训练中的弹着点分布情况, 以采取针对性训练手段提高训练者的射击水平.首先介绍了自动报靶的发展历程, 之后对现有的弹孔识别方法进行了分析研究, 随后提出了使用滞后阈值边缘检测算法能够更好地识别弹孔.首先对前后两张图像进行仿射变换, 实现图像校正, 对校正之后的两张图像进行滞后阈值边缘检测使得弹孔信息更加明显, 通过分析弹孔图像, 发现相邻两张图像唯一的不同点就是弹孔区域, 采用简单快速的减影运算完成对弹孔区域的获取, 最后通过开运算消除弹孔以外干扰区域.实验结果表明, 所提出的算法识别弹孔准确度高, 运行时长符合实弹训练要求.%In the shooting training of outdoor, visual inspection technology is used to report target automatically, which can not only prevent training safety accidents, but also obtain the distribution of bullet points in the trainers' shooting training, so as to improve the shooting level of trainers by adopting targeted training methods.Firstly, we introduce the development of automatic target reporting, then analyze and study the existing bullet identification methods, and then propose hysteresis threshold edge detection algorithm to better identify the bullet holes.First of all, two images are carried out affine transform to achieve correction.The edge detection of the hysteresis threshold of two images after correction makes the bullet hole information more obvious.Through the analysis of the bullet hole image, it is found that the only difference between the two adjacent images is the region of the bullet hole.So a simple and fast subtraction operation is used to obtain the hole area.Finally, the interference area outside the hole is eliminated by open operation.The experiment shows that the algorithm has high accuracy in identifying bullet holes and the running time is consistent with the training requirements.
展开▼
