Research On Target Recognition Based On Vision For Mobile Robot

Target Recognition Based on Vision for Mobile Robot is a research hotspot in the field of mobile robot research.It has a broad application prospect and has important theoretical significance and application value.Its core problem is to improve the accuracy,speed and robustness of the target recognition algorithm.Based on the analysis of existing target recognition algorithms at home and abroad,this paper makes an in-depth study on the target recognition problem of vision-based mobile robot systems,mainly involving target feature extraction and feature matching algorithms.The main tasks completed are as follows:(1)An improved image matching algorithm based on PCA-SIFT and RANSAC is proposed.Firstly,the SIFT features of the image are extracted.Secondly,the principal component analysis method is used to reduce the dimension of the SIFT feature descriptor,which is reduced from 128-dimensional to 20-dimensional,and then the Euclidean distance is used for feature matching.Finally,the traditional RANSAC algorithm is improved.Feature point matching for error correction.Experimental results show that the algorithm improves the efficiency of image matching.(2)An image matching algorithm based on Mahalanobis distance and PCA-SIFT is proposed.Firstly,the SIFT features of the image are extracted,then the principal component analysis method is used to reduce the dimension of the SIFT feature descriptor.Finally,the two-dimensional Mahalanobis distance matching algorithm is used for feature matching.Although the Mahalanobis distance is more complex than the Euclidean distance,the Mahalanobis distance is not affected by the correlation between the dimension and the eigenvector,and has good affine invariance,making the eigenvector more representative.Experimental results show that the algorithm improves the accuracy of image matching.