摘要
在一般图像处理步骤的基础上,屏弃了轮廓成形的常规路线,对锐化处理步骤进行了特殊处理得到了车框货物的顶层图像。提出了一种基于邻域像素统计的边界点查找算法,对货物顶层图像中心的提取行之有效。此项技术为实现运输车辆自动定位的功能解决了关键问题。
This paper abandons profile figuration method based on normal processes for the image recognition and the top layer products image is extracted after using special image peaking method. In addition, an arithmetic for finding verge points based on the neighborhoods is put forward, which is proved feasible in practice. This technology resolves the key problem in the automatic truck frame fixed position.
出处
《机械制造与自动化》
2013年第2期137-140,共4页
Machine Building & Automation
关键词
图像识别
车框定位
梯度锐化
邻域查找
image recognition
truck frame fixed position
grads peak
neighborhoods findi searching