Recognition Method Based on Gabor Wavelet Transform and Discrete Cosine Transform

A fingerprint recognition method based on Gabor wavelet transform and discrete cosine transform (DCT) was proposed. The multichannel Gabor filters and improved binarization algorithm are adopted to enhance the fingerprint images. The hybrid strategy of the discrete cosine transform method and the Gabor wavelet transform algorithm was proposed to form an improved Gabor wavelet transform applied to the fingerprint feature extraction and feature reduction. Finally, the nearest neighbor distance detector and Euclidean distance are used as classifier to classify the fingerprint. The results of simulation experiments prove that the proposed strategy can effectively shorten the recognition time and maintain a considerable recognition rate.