Integrated Shuffled Singular Value Decomposition and Wavelet Difference Reduction based Image Compression

The paper presents a new lossy compression technique in which compression is achieved by combining two compression based algorithms. The first method is wavelet difference reduction (WDR) based image compression and the other is shuffled singular value decomposition (SSVD) based image compression. Both the methods are integrated in a way to achieve high compression with better image quality. WDR is a compression technique that involves applying DWT with differential reduction encoding. Shuffled Singular Value Decomposition (SSVD) is enhanced version of singular value decomposition (SVD) that gives high PSNR values for same rank. DWT is applied on the image to be compressed at the encoder side and then WDR based compression is applied on low frequency sub-band. SSVD based compression is applied on high frequency sub-bands parallely. At the decoder side image quality enhancement is carried out. The proposed method is compared with other methods. It is seen that psnr values of proposed method are greater than other methods for same amount of compression.