Low-resolution face recognition methods based on global and local feature integration
Due to the blurred face features in low resolution images,the accuracy of face recognition results is low,and the low resolution face recognition method based on global and local feature integration is proposed.Principal component analysis method to extract the global features of low resolution face image,through the LBP algorithm of local features,build a convolution neural network containing feature integration module and classification recognition module,input extraction of global and local features,after learning output low resolution face recognition results.The experimental results show that the accuracy of the low-resolution face recognition results under the design method is up to 94%,and the proposed method is effective and superior.
global featureslocal featuresfeature integrationlow resolutionface recognition
NETL
NSTL
万方数据
