An Online Surface Defects Detection System for Step-Axis

Abstract:

Surface defects detection is an important application of machine vision. In this paper, an online surface defects detection system of step-axis is studied based on image recognition. To ensure the real-time property, a fast axial surface defects inspection method is put forward, including improved median filtering to reduce noise, gray variation for fast judgment, the maximum variance method (OTSU) to select threshold automatically, contour features for feature extraction, mathematical morphology to detect defect targets, and finally, support vector machine (SVM) to classify and recognize the surface defects of ladder shaft. Experimental results show that the system can detect surface defects of the step-axis in 0.5s, which can meet the real-time requirements.

• Periodical: Advanced Materials Research (Volumes 472-475)
• Pages: 954-957
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.472-475.954
• Online since: February 2012
• Authors: Ji Li Lu, Ming Xing Lin
• Keywords: Defect Detection, Feature Extraction, Gray Variation, Median Filter, Support Vector Machine (SVM)

References

[1] Shilei Li: Hot Working Technology. Vol.39(2010) ,p.188

[2] Gang Yab, Xiaodong Xu: J. Optoelectronics ·Laser. Vol.17(2006),p.107

[3] Bin Wu,Longtao Li, Xiuyan Wang: Engineering Mechanics .Vol.20(2003),p.149

[4] Shenghua Ye, Jigui Zhu, Zhong Wang, Xueyou Yang: Engineering Sciences. Vol.1(1999), p.49

[5] Fuguo Dong,Hui Fan, Da Yuan I.J. Digital Content & its Applications. Vol.4(2010) p.79

[6] Zunbing Sheng, Xianyu Cui,Guoan Gao: J. Shanghai Jiaotong Univ. Vol.41(2007）p.911

[7] Kefei Cheng,Cong Zhang:Computer Simulation.Vol.23(2006）p.92

[8] Vanpnik V.:Statistical learning theory(Wiley, New York 1998)