摘要
在线膛炮身管内膛疵病图像识别中,由于阴阳膛线之间的高度差使得二者的过渡区形成阴影,拍摄得到的图像呈现灰度不连续的交替条纹,给疵病图像的分析带来困难,因此,需要对图像进行处理,使得处理后图像灰度连续。采用对膛线过渡区图像进行全局阈值分割、移除小目标、膛线过渡区灰度填充的方法实现对膛线过渡区的有效处理。
Owing to the altitude difference between grooves and lands brought on shadow of the transition region and the shooting image appears as gray discontinuity of alternate stripes, image analysis of flaw is difficulty in image recognition of artillery barrel, thereby, the image needs to be processed and image gray becomes continuous. The image of rifling transition region is processed effectively by global threshold segmentation, removing small target, graying filled in rifling transition region in this paper.
出处
《火力与指挥控制》
CSCD
北大核心
2014年第3期163-164,共2页
Fire Control & Command Control
关键词
火炮身管
膛线
图像处理
artillery barrel, rifling, image processing