Research on Positioning Datum Recognition Based on Image Processing and SVM
In order to solve the problem of classification of multi-feature positioning datum in the process of robot automatic hole making,this paper proposes a positioning datum classification method based on image processing and SVM.The image containing multi-feature positioning datum is segmsegmed by reference ROI region based on glob-al optimal threshold,and the statistical variance,roundness and gray mean of feature region of the reference feature are calculated.It is used as the characteristic parameter to describe the positioning datum and combined with SVM to realize the identification and classification of the positioning datum.The results show that the accuracy of the method is 94.3％,the recall rate is 95.9％,and it has certain application value for the whole process automation of robot automatic drilling.
digital image processingsupport vector machinesclassifiersfeature vectors surface
万方数据
维普
