基于特征融合的堆叠工件分类识别研究
首发时间:2019-05-29
摘要:本文针对堆叠工件的识别问题,提出了一种基于决策融合的方法,旨在提高目标工件的识别准确率。使用SVM支持向量机作为分类器,提取工件的LBP特征与HOG特征且将其用于训练分类器,对不同情况下的特征分类器进行对比分析,分别从两种特征分类器中选择分类准确率较高的分类器进行融合,得出分类准确率最高的分类器。最后,将LBP特征分类器,HOG特征分类器与融合后的特征分类器分别做样本识别实验,对实验结果进行对比分析,证明了本文算法能够达到更高的分类准确率。
For information in English, please click here
Research on Classification and Recognition of Stacked Workpieces Based on Feature Fusion
Abstract: In this paper, a method based on decision fusion is proposed for the identification of stacked workpiece, which aims to improve the recognition accuracy of target workpiece. SVM support vector machine (SVM) is used as a classifier to extract the HOG features and LBP features of the workpiece and use them to train the classifiers. The feature classifiers in different cases are compared and analyzed. From the two feature classifiers, the classifiers with higher classification accuracy are selected for fusion, and the classifiers with the highest classification accuracy are obtained. Finally, the LBP feature classifier, the HOG feature classifier and the fused feature classifier are respectively used in the sample recognition experiment, and the experimental results are compared and analyzed. It is proved that the proposed algorithm can achieve higher classification accuracy.
Keywords: decision fusion LBP feature HOG feature
引用
No.****
动态公开评议
共计0人参与
勘误表
基于特征融合的堆叠工件分类识别研究
评论
全部评论0/1000