ABS齿圈环形表面缺陷检测方法
发布时间：2019-06-02 12:12
【摘要】：针对传统人工方式检测汽车用ABS齿圈环形表面缺陷时存在检测效率低、易错检漏检的问题,提出一种基于图像处理的ABS齿圈环形表面缺陷检测方法。根据生产实际设计并组装ABS齿圈环形表面缺陷在线视觉检测系统,利用旋转电缸结合齿圈托台带动齿圈旋转,由线阵CCD扫描并得到齿圈环形表面图像,在经过基于OpenCV编写的图像处理算法处理后根据缺陷所在区域判断缺陷类型,进而判断齿圈合格性。通过实验将系统检测与人工检测结果进行对比,结果表明,每个齿圈平均检测时间≤4 s,缺陷分类正确率≥92%。
[Abstract]:In order to solve the problems of low detection efficiency and easy error detection when the traditional manual method is used to detect the annular surface defects of ABS ring for automobile, a method of detecting annular surface defects of ABS tooth ring based on image processing is proposed. According to the production practice, the on-line visual detection system of ABS ring surface defect is designed and assembled. The rotating electric cylinder combined with the tooth ring bracket is used to drive the tooth ring to rotate, and the ring surface image of the tooth ring is scanned and obtained by linear CCD. After the image processing algorithm based on OpenCV, the defect type is judged according to the defect area, and then the tooth ring qualification is judged. The results of system detection and manual detection are compared through experiments. The results show that the average detection time of each tooth ring is less than 4 s, and the correct rate of defect classification is 鈮，
本文编号：2491090
[Abstract]:In order to solve the problems of low detection efficiency and easy error detection when the traditional manual method is used to detect the annular surface defects of ABS ring for automobile, a method of detecting annular surface defects of ABS tooth ring based on image processing is proposed. According to the production practice, the on-line visual detection system of ABS ring surface defect is designed and assembled. The rotating electric cylinder combined with the tooth ring bracket is used to drive the tooth ring to rotate, and the ring surface image of the tooth ring is scanned and obtained by linear CCD. After the image processing algorithm based on OpenCV, the defect type is judged according to the defect area, and then the tooth ring qualification is judged. The results of system detection and manual detection are compared through experiments. The results show that the average detection time of each tooth ring is less than 4 s, and the correct rate of defect classification is 鈮，
本文编号：2491090
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2491090.html
