基于目标特征的机械零件表面缺陷检测方法.PDF

2016 年 软 件 2016, Vol. 37, No. 4 第 37 卷 第 4 期 COMPUTER ENGINEERING SOFTWARE 国际 IT 传媒品牌 基于目标特征的机械零件表面缺陷检测方法 王诚诚，李文森，雷 鸣，管声启 (西安工程大学 机电工程学院，陕西 西安 710048) 摘 要: 针对机械零件表面缺陷的特点，提出了一种基于目标特征的检测方法。首先，通过对测试图像进行均值 滤波，以消除噪声对检测的影响；其次，通过分析表面缺陷特征设置分割阈值，在此基础上，实现对检测目标分割； 最后，采用形态学滤波消除噪声和孤立点对检测结果的影响。实验表明，本文采用的方法能够有效抑制图像背景干扰， 能够有效的实现机械零件表面缺陷准确检测。 关键词: 机械零件；表面缺陷；目标特征；检测方法 中图分类号: TP391.41 文献标识码: A DOI： 10.3969/j.issn.1003-6970.2016.04.009 本文著录格式：王诚诚，李文森，雷 鸣，等. 基于目标特征的机械零件表面缺陷检测方法[J]. 软件，2016，37（4）：3234 Surface Defect Detection Method of Mechanical Parts Based on Target Feature WANG Cheng-cheng, LI Wen-sen, LEI Ming, GUAN Sheng-qi (College of mechanical electronic Engineering, Xi’an Polytechnic University, Xi’an, 710048, China) 【Abstract 】: Through the analysis of surface defect characteristic about mechanical parts, a new method of based on target feature was presented to surface defect detection of mechanical parts. First of all, the test image is processed by mean-filter in order to eliminate the effect of niose on the detection. Secondly, based on the analysis of the characteristic of the surface defect, the segmentation threshold is set, and segment surface defect of mechanical parts. Finally, the ef- fect of noise and outlier on the detection result was eliminated by morphological filtering. Experiment shows that the method can effectively suppress the image background interference, and more accurately realize the surface defect de- tection of mechanical parts. 【Key words 】: Mechanical parts; Surface defect; Target feature; Detection method 0 引言 等[7] 。其中，零均值化方法是通过构造零均值化图， 并采用阀值分割出缺陷区域，这种算法虽然简单，但 在机械零件加工以及使用过程中，机械零