基于图像处理技术, 采用嵌入式机器视觉处理系统, 设计了一种用于继电器触点间距的智能检测装置和方法。在硬件方面, 以Cortex-A8处理器为中心, 将控制、处理与测量于一体, 并采用红外光源作为照明系统, 抑制金属触点的光发射; 在算法方面, 以继电器触头静触片的接触点图像区域为特征图像, 并采用多分辨率金字塔分层匹配技术, 实现对检测对象的准确识别和定位, 同时根据继电器内部组件位置的对应关系, 确定继电器触点边缘检测区; 在检测区内, 先采用像素级的边缘检测算法确定两接触点边缘后, 再利用亚像素边缘检测技术, 实现边缘的精确定位, 并检测出继电器触点间的最短距离。本文所提出的方法对继电器质量检测和性能评估具有重要的意义。

Based on the image processing technology and the embedded mechanic visual processing system, we have designed an equipment and method to intelligently detect the relay contact spacing. In hardware respect, based on the Cortex-A8 processor as the system center, the controlling circuit of imaging system is integrated with image processing and measuring, and at the same time it takes advantage of the infrared light source as an illuminating system to suppress the luminous emission of the metal contacts. In the algorithm respect, the rapid identification of detecting object can be realized by taking the image region of the static contact point pad of relay as the feature image and by adopting multi-resolution hierarchical matching technique based on pyramid algorithm. According to the corresponding position relationship between internal components of relay, the relay contact edge detection area can be locked, and after determining the edge of the two contact by using the pixel-level edge detection algorithm, we can accurately locate the detected edge by sub-pixel edge detection technology, and finally successfully detect the minimum distance between the relay contacts. The method proposed in this paper is of great value for the quality evolution and detection of the relay.