China Mechanical Engineering

Previous Articles     Next Articles

Defect Detection Method of Linear Guide Surfaces Based on HSI Color Space

ZHOU Youhang;MA Zhuxi;SHI Xuanwei;KONG Tuo   

  1. School of Mechanical Engineering,Xiangtan University,Xiangtan,Hunan,411105
  • Online:2019-09-25 Published:2019-09-24

HSI颜色空间下的直线导轨表面缺陷检测方法

周友行;马逐曦;石弦韦;孔拓   

  1. 湘潭大学机械工程学院,湘潭,411105
  • 基金资助:
    国家自然科学基金资助项目(51775468,51375419)

Abstract: In order to solve the problems that tiny defects of linear guide surfaces were not accurately detected due to the effects of background texture, a visual detection method was proposed for surface defects of linear guides in HSI color space. Firstly, the images of the surfaces were converted from RGB space to HSI space for obtaining the components of hue,saturation and brightness. Then, the dimensions of images were reduced, which was based on the cumulative contribution rate of PCA, and a mixed gray model was established by the low-dimension images.After that,the corresponding weighting coefficients were optimized by PSO algorithm, and finally, the defects were segmented by threshold. The results of experiments and calculations show that the proposed method may accurately detect 4 types of defects compared with the extraction method in RGB space.

Key words: linear guide, defect detection, HSI color space, principal component analysis(PCA), particle swarm optimization(PSO) algorithm

摘要: 为解决直线导轨表面微小缺陷受背景纹理影响、无法准确检测的问题,提出了一种HSI颜色空间下的直线导轨表面缺陷视觉检测方法。将直线导轨表面图像由RGB空间转换到HSI空间,得到色调、饱和度和亮度的分量图;采用主成分分析法对各分量图像进行降维,建立混合灰度模型;利用粒子群算法优化其相应的加权系数;通过阈值分割完成缺陷检测。实验及计算结果表明,相比RGB空间的缺陷提取,该方法能准确检测出常见的直线导轨表面四种类型缺陷。

关键词: 直线导轨, 缺陷检测, HSI颜色空间, 主成分分析, 粒子群优化算法

CLC Number: