IMAGE COMPRESSION USING CDF 5/3 LIFT DWT AND SVM

Revathy K.R, Arun Vikas Singh

Abstract: Image compression has become an unavoidable process in multimedia and communication applications .Here this paper present a better method of image compression using based DWT and Support Vector Machines(SVM). The Discrete Wavelet Transform(DWT) used here is 5/3 lift based DWT. After the application of DWT the Support Vector Machine will classify the minimum value wavelet coefficients. The compression can be done by combining Support Vectors(SVs) and corresponding weights, then the coefficients are quantized and encoded. By doing the reverse process the reconstructed image is generated. Finally the results shows that the proposed method provides high quality image with good Peak Signal to Noise Ratio(PSNR) and Compression Ratio(CR).

Keywords: Discrete Wavelet Transform(DWT),Support Vector Machine(SVM), PSNR, Compression Ratio(CR)

DOI: https://doi.org/10.15623/ijret.2016.0516050