基于机器视觉的阴极铜表面质量检测系统研究

为了解决了目前人工对于阴极铜表面质量检测存在的问题,实现阴极铜的全自动化检测与筛选。本文将图像处理技术应用于阴极铜表面质量的检测。论文的研究方法如下:首先选择试验需要的硬件和软件,然后通过试验采集照片,通过系统获得数据,最后把实际测量的数据和系统数据对比,分析之后得出结论。表面检测系统的流程设置为:图像预处理、相机标定、立体较正、立体匹配、数据计算。应用的原理和方法是双目视差原理、三角测量法、逐渐像提取法、以及Sobel算子和BM算法。把检测系统计算出相关数据与试验实际测量得出的数据进行对比,发现两种数据差别极小,检测结果一致,验证了检测系统的正确性,说明该系统能够满足实际生产的需求。最终建立一套阴极铜表面质量检测系统,确立了系统的结构设计和位置安装。本系统解决了当前人工筛选存在的问题,实现了阴极铜的全自动化生产。

Research on surface quality inspection and sorting system of cathode copper based on machine vision

In order to solve the problems of current artificial screening,realize the automatic detection and screening of cathode copper.In this paper,the application that image processing technology is used in the detection of surface quality of cathode copper. The research methods of this paper are as follows: First of all, according to the actual production conditions, the parameters of the hardware devices such as light source, camera, lens, image acquisition card and so on are determined; Then, the machine vision technology is studied. Based on the Halcon platform, the recognition algorithm based on edge detection and threshold segmentation is used to analyze, comprehend and extracted the required information.Finally, the related experiments are completed, and the problems are found, which provides the basis for further optimization.In this paper, two kinds of image recognition algorithms based on edge detection and threshold segmentation are studied mainly.Through experiments, the advantages and disadvantages of the two algorithms are compared respectively from three aspects: error detection rate,missing rate and processing time. The average of error detection rate of A (image recognition algorithm based on edge detection) algorithm was 3.3% for 100 cathodes of copper cathode, and 3.7% for B (image recognition algorithm based on threshold segmentation ). The average of solution A the missing rate was 11.5%,the average missing rate was 12.4% for Plan B, the processing time was 0.40 s for Plan A, and 0.75 s for Plan B.Based on the above analysis, the image recognition algorithm based on edge detection is chosen to complete the actual production needs.Finally, a set of system that can detcte the surface quality of cathode copper was established , the design of structure and installation of location about this system was set up. In this system, the problems of artificial screening has been solved , full automatic production of cathode copper has been achieved.