A License Plate Super-resolution Reconstruction Algorithm Based on Manifold Learning

To address the problem of low resolution in surveillance video, which leads to difficulty in recognizing license plates, this paper presents a new license plate image super-resolution reconstruction method based on manifold learning. Firstly, the mapping between low-resolution images and high-resolution images in training set is obtained by learning method. Then image feature vectors are extracted by linear discriminant analysis (LDA) algorithm and its parameters are modeled by locally linear embedding (LLE) algorithm. Finally the high resolution image is reconstructed by the mapping relation. Experimental results show that the proposed algorithm has better effect on super-resolution restoration for real low-resolution plate image, significantly improve the license plate character identification.