基于机器视觉的轴类零件识别  

Recognition of Shaft Parts Based on Machine Vision

作　　者：安爱琴[1] 马玉峰[1] 张谦[1] 

机构地区：[1]河南科技学院机电学院,新乡453003

出　　处：《煤矿机械》2015年第12期292-294,共3页Coal Mine Machinery

基　　金：河南科技学院机电学院青年基金项目;河南科技学院大学生创新项目

摘　　要：采用机器视觉手段进行轴类零件识别。通过采集待识别零件样本信息,利用MATLAB软件编程实现对样本的图像处理和轮廓提取,获得特征量信息。然后通过图像标定和亚像素最小二乘法进行直线拟合获得轴的边缘轨迹,完成轴的直径和同轴度识别。进行视觉识别结果和人工测量结果比较,结果表明,采用机器视觉进行轴类零件识别,相对误差为0.001,检测精度具有一定的可靠性,且检测速度快,便于实现批量化、在线实时检测,具有广阔的应用前景。The paper identified the shaft, with the machine vision means. Based on the collected samples of the parts, image processer, contour extraction and feature quantity information were executed, using MATLAB software. Then straight lines using the least squares method were fitted based on the edge path axes, the shaft diameter and concentricity were completed. The results of visual identification and manual measurement were compared. The results showed that the relative error of the machine vision recognition was 0.001, which indicated that the detection accuracy was reliability,speed, and easy to implement batch, on-line real-time detection. The conclude had broad application prospects.

关 键 词：机器视觉 轴类零件 最小二乘法拟合 MATLAB软件 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]