Optical Footprint Image Recognition Algorithm Based on Metric Learning and SVM

Footprint recognition plays an important role in criminal investigation, security protection and other fields. In terms of the key problems of feature extraction and classification, a footprint recognition algorithm is proposed based on optical footprint images. The algorithm first transforms the footprint image into the complex frequency domain through the dual-tree complex wavelet transform (DTCWT), then extracts the histogram of oriented gradient (HOG) features of the footprint image in the complex frequency domain. Finally, a symmetric projection matrix is obtained through metric learning. The symmetric projection matrix is used to project the original features to a new feature space. And then the SVM is used to identify the optical footprint in the new feature space. The experimental results show that on the optical footprint images of 134 people, the recognition accuracy of the algorithm in this paper is improved, which is valid to help the application and research of footprints.