摘要
基于机器视觉的零件检测系统由于具有非接触、实时性强、速度快等优点广泛应用于各种工业生产中,提出了一种基于边缘跟踪的零件缺陷边缘智能检测算法,很好的检测到了完整的缺陷边缘,为特征提取提供了高质量的缺陷边缘参数。采用基于支持向量机的分类识别算法,避免了神经网络算法中需要多样本和过度拟合的问题,通过对比分析选择合适于本系统的核函数,并运用基于交叉验证和网格搜索的参数选择方法找到核函数的最佳参数,采用一对一的投票策略进行分类训练和测试,最后对采集到的缺陷零件样本进行了分类测试实验,达到预定的较高的检测精度。
With the characters of non-contact, real-time and fast-speed, the detecting system of machinery parts based on machine vision has been applied in industry production practice widely. This paper propose a intelligent algorithm of detecting defect edge based on edge tracking, which can detect the defect edge integrally and provide the parameters of defect edge for feature extraction. This paper use the classification algorithm of support vector machine （SVM）, which avoided the questions that neural network algorithm needs more samples and easily to overfitting. This paper chooses the appropriate kernel function,and gets the optimal parameter of kernel function with the method of cross validation and grid search parameters selection, uses one-to-one voting strategy to train and test classification sample. By testing the classification of the collected defective parts sample,this system achieves to the high detecting precision.
出处
《电子测量技术》
2012年第1期80-84,共5页
Electronic Measurement Technology
关键词
支持向量机
机器视觉
图像处理
缺陷检测
SVM （support vector machine） machine vision image processing defect inspection
