STUDY OF INCREASE IN GLOBAL COMPONENTS IN HYBRID WAVELETS ON DATA COMPRESSION

Abstract

This paper presents a hybrid wavelet transform technique which studies the effect of global components on the quality of image compression. Hybrid wavelet transform is generated using two different component orthogonal transforms. One orthogonal transform represents global featuresof image in betterway and another is used to represent local features. Walsh transform of size 8x8 is used as a base transform i.e. to represent global characteristics of image. Other transforms like DCT, Discrete Real Fourier Transform,DiscreteHartley transform (DHT), Discrete Sine Transform (DST), Discrete Kekre Transform (DKT) and Slant transform of size 32x32 are used to focus on local characteristics of an image.256x256 hybrid wavelet transform is generated and multiple iterations of global components are included using columns of base transform and its effect on reconstructed image quality is observed in terms of Root Mean Square Error (RMSE) and Peak Signal to Noise Ratio (PSNR).  From the experiments it has been observed that when DCT is used to extract local features, best results are obtained among all combinations with Walsh transform. These results are also compared with Walsh transform and observed to be much superior at higher compression ratios giving better image quality.