Abstract:

According to the energy aggregation properties of image wavelet coefficients, a new wavelet image compression algorithm based on matrix completion is proposed. On the basis of singlelayer discrete wavelet transform, the algorithm takes sparse samples from high frequency sub-bands, followed by entropy coding for image compression. For decompression, the singular value shrinkage iteration algorithm is adopted to implement matrix completion, recovering high frequency sub-bands from the sampled coefficients of small amount, then the image is reconstructed via inverse wavelet transform. Experimental results show that the proposed algorithm is simple, and has a better compression performance when the encoding bit rate is greater than 1 bit/pixel.