An Efficient Compression Algorithm for Finger Images Recognition System
Jia Wang · Tv Engineering · 2005
Aiming at coding finger images efficiently an improved image coding algorithm based on quadtree classification and trellis coded quantization is proposed in this paper. The coefficients in the high frequency sub-images of all scales are transformed with 2×2 DCT and reordered to make the energy of these sub-images concentrate in the low frequencies of the corresponding subbands. Then the DCT transformed and reordered coefficients are encoded with quadtree classification and TCQ. Simulation results show that, compared to WSQ and JPEG2000, this algorithm can gain better compression performance.