A Novel Image Compression Algorithm Using the Second Generation of Curvelet Transform and SVM

This paper proposes a novel image compression algorithm which combines support vector machine (SVM) regression with curvelet transform. Firstly, the original image was decomposed into curvelet coefficients using fast discrete curvelet transform. Then different scales of quantized curvelet coefficients were selected for lossy compression and entropy encoding. For example, the lowest subband is encoded by differential pulse code modulation (DPCM) for including most image energy. The finer scale subbands are compressed by SVM regression, which approximates the curvelet coefficients using a fewer support vectors and weights. And some of the finer scale subbands are discarded directly because they only contain a little amount of energy and have little noticeable effect on the image quality. Compared with image compression method based on wavelet transform, experimental results show that the compression performance of our method gains much improvement. Moreover, the algorithm works fairly well for declining block effect at higher compression ratios.