锻压技术

基于机器视觉的弯曲角度检测

英文标题：Bending angle detection based on machine vision

作者：吴雅莎肖小亭陈康徐信

单位：广东工业大学

分类号：TG385.9

出版年,卷(期):页码：2018,43(8):184-189

摘要：

针对全自动冲压生产线中弯曲回弹角度人工检测困难的问题，提出一种基于机器视觉的检测方法。介绍了弯曲角度检测系统的设计方案，通过相机标定确定相机的内外参数，用以校正采集图像提高检测准确度，基于Halcon视觉算法对图像进行处理，识别出弯曲角度两边目标线段，计算由每个弯曲角的内外边界所组成的内角和外角的平均值，最终得到检测弯曲角度。其中，图像处理算法包括：BLOB分析将弯曲件从背景中分离出来；形态学处理得到弯曲件的内外边界；提取亚像素边缘；直线段的分割和拟合得到检测目标。该方法实现了非接触式亚像素精度检测弯曲角度，为弯曲件的质量检测提供判断依据。

For the problem of difficult manual detection of bending springback angle in automatic stamping production line, a detection method based on machine vision was proposed. Then, the bending angle detection system was introduced, and the internal and external parameters of camera were calibrated by the camera to correct the image and improve the detection accuracy. Based on the vision algorithm Halcon, the image was processed, and the target line segments on both sides of bending angle were identified. Furthermore, the average values of the internal and external angles composed of inner and outer boundaries for each bending angle was calculated, and the bending angle was detected at last. Moreover, the bending part was separated from the background by image processing algorithm including BLOB analysis, and the inner and outer boundaries of bending parts were obtained by morphological processing. Finally, the sub-pixel edges were extracted, and the detection targets were obtained by the segmentation and fitting of line segments. The method can realize non-contact sub-pixel precision detection of bending angle and provid the judgment basis for quality detection of bending parts.

基金项目：

广东省战略性新兴产业发展专项(2012A090100014)；广东省教育部产学研结合项目(201213091100024)；广州市科技项目(2014J2200084)

吴雅莎（1992-），女，硕士研究生，E-mail：yasha2008@163.com；通讯作者：肖小亭（1957-），男，博士，教授，博士生导师，E-mail：xiaoxt@gdut.edu.cn

参考文献：

［1］徐信，肖小亭，温锡昌，等. 拉深件缺陷在线检测系统设计
［J］. 锻压技术. 2016, 41(7): 97-101.

Xu X, Xiao X T, Wen X C, et al. Design of on-line detection system for drawing defects
［J］. Forging & Stamping Technology, 2016, 41(7):97-101.

［2］陈琦，阮鸿雁. 基于机器视觉的滑动轴承缺陷检测系统设计
［J］. 组合机床与自动化加工技术，2017，(5): 92-95.

Chen Q, Ruan H Y. Design of detection system for slide beating defects based on machine vision
［J］.Modular Machine Tool & Automatic Manufacturing Technique，2017，(5): 92-95.

［3］Muniraja I, Guo C, Malallah R, et al.Subpixel based defocused points removal in photon-limited volumetricdataset
［J］. Optics Communications, 2017, 387:196-201.

［4］Carsten Steger, Markus Ulrich, Christian Wiedemann. Machine Vision Algorithms and Applications
［M］.Translated by Yang S R, Wu D J, Duan D S, et al. The First Edition. Beijing: Tsinghua University Press, 2008.

［5］朱云芳.摄像机径向畸变校正和内参估计的单图标定方法
［J］.光电工程,2012,39(9):125-131.

Zhu Y F. Calibration of radial distortion and intrinsic parameters using a single image
［J］.Opto-Electronic Engineering,2012, 39(9):125-131.

［6］贾丹.摄像机现场标定算法研究
［D］. 哈尔滨：哈尔滨工程大学, 2007.

Jia D. Researches on Site Camera Calibration Methods
［D］. Harbin：Harbin Engineering University, 2007.

［7］王辉，李晋光.基于BLOB的图像查询系统应用研究
［J］.电子设计工程,2011,19(16):51-54.

Wang H, Li J G. Research on the image query system based on BLOB
［J］.Electronic Design Engineering,2011,19(16):51-54.

［8］Guo Y Y, Guo Q. A sort of gray-scale morphological dilations and erosions
［J］. Chinese Quarterly Journal of Mathematics, 2017, 32(2):181-186.

［9］Canny J. A computational approach to edge detection
［J］.IEEE Transactions on Pattern Analysis and Machine Intelligence, 1986,8(6):679-698.

［10］ Ramer U. An iterative procedure for the polygonal approximation of plane curves
［J］. Computer Graphics and Image Processing, 1971, 1:244-256.

服务与反馈：

【文章下载】【加入收藏】