Brake pad image classification algorithm basedon color segmentation and information entropy weighted feature matching
ZHAO Lei1, ZHANG Wen1, SUN Zhenguo1,2, CHEN Qiang1,2
1. Key Laboratory for Advanced Materials Processing Technology of Ministry of Education, Department of Mechanical Engineering, Tsinghua University, Beijing 100084, China; 2. Yangtze Delta Region Institute of Tsinghua University, Jiaxing 314006, China
Abstract：In appearance inspection of car brake pads based on machine vision, the segmentation of the brake pad image and the extraction and matching of the shape features are keys of brake pads classification. In order to realize high-precision shape classification, this paper proposes a brake pad image classification algorithm including foreground segmentation, geometric feature selection and template matching. First the RGB image captured by the camera is converted to HSV color space and using saturation channel the brake pad is segmented from the dark belt background. Then the multi-dimensional geometric features of the area are extracted. Finally, an improved feature matching algorithm based on information entropy weighting is proposed, in which Manhattan distance of the feature space is weighted by the information entropy of the brake pad features. The experimental results show that the accuracy of the algorithm is 95.00%, and the average processing time is 110 ms. It can be applied to the real-time automatic classification procedure during brake manufacturing.
[1] 左东祥, 陈晓荣. 基于HALCON的刹车片尺寸和表面缺陷检测系统[J]. 电子科技, 2016, 29(11):78-80. ZUO D X, CHEN X R. Detection system of brake based on HALCON[J]. Electronic Science and Technology, 2016, 29(11):78-80. (in Chinese) [2] 马艳宁, 陈晓荣, 张运涛. 基于Halcon的刹车片字符检测算法研究[J]. 电子科技, 2016, 29(10):101-103. MA Y N, CHEN X R, ZHANG Y T. Research on the brake character detection algorithm based on Halcon[J]. Electronic Science and Technology, 2016, 29(10):101-103. (in Chinese) [3] 曾强, 孙坚. 基于HALCON的刹车片轮廓缺陷检测方法研究[J]. 煤矿机械, 2016, 37(8):143-146. ZENG Q, SUN J. Research on contour defects detection for break pad based on HALCON[J]. Coal Mine Machinery, 2016, 37(8):143-146. (in Chinese) [4] BELONGIE S, MALIK J, PUZICHA J. Shape matching and object recognition using shape contexts[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2002, 24(4):509-522. [5] MOKHTARIAN F, ABBASI S, KITTLER J. Efficient and robust retrieval by shape content through curvature scale space[C]//Proceedings of the International Workshop on Image Databases and Multi-Media Search. Amsterdam, Netherlands:IAPR, 1996:35-42. [6] CHUI H, RANGARAJAN A. A new algorithm for non-rigid point matching[C]//Proceedings of the 2000 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Hilton Head Island, USA:IEEE, 2000:44-51. [7] ADAMEK T, O'CONNOR N E. A multiscale representation method for nonrigid shapes with a single closed contour[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2004, 14(5):742-753. [8] DU P J, TAN K, SU H J. Feature extraction for target identification and image classification of OMIS hyperspectral image[J]. Mining Science and Technology, 2009, 19(6):835-841. [9] 冯彦铭, 谢家龙, 李林升, 等. 基于八邻域跟踪算法的封闭区域几何尺寸测量[J]. 机械工程与自动化, 2013(6):170-171, 173. FENG Y M, XIE J L, LI L S, et al. Closed area geometric measurement based on eight neighborhood tracking algorithm[J]. Mechanical Engineering & Automation, 2013(6):170-171,173. (in Chinese) [10] HU M K. Visual pattern recognition by moment invariants[J]. Information Theory, IRE Transactions on, 1962, 8(2):179-187. [11] 肖汉光, 蔡从中. 特征向量的归一化比较性研究[J]. 计算机工程与应用, 2009, 45(22):117-119. XIAO H G, CAI C Z. Comparison study of normalization of feature vector[J]. Computer Engineering and Applications, 2009, 45(22):117-119. (in Chinese) [12] SHANNON C E, WEAVER W. The mathematical theory of communication[J]. Physics Today, 1950, 3(9):31-32.
2018, Vol. 58 
