Complex Wavelet Based Image Analysis

The main goal behind this thesis is estimating the recovered image from the distorted or corrupted image. The advantages of Complex Wavelet Transform and Hyperanalytic Wavelet Transform over real standard wavelet transforms provides more scope in the areas of image denoising and image compression. The other objective of the present work, is extending the DWT implementation to Diversity Enhanced DWT. A version of Hyperanalytic Wavelet Transform (HWT) is implemented with a zero order wiener filtering for image analysis. The proposed HWT based method allows the usage of multi wavelets compared to Complex Wavelet Transform( DTCWT) where a set of predefined analysis and synthesis filters are used. Another objective of the work is to analyze how the Dual-Tree Complex Wavelet Transform can be applied with huffman coding in image compression. The performance evaluation of different algorithms is based on metrics like Peak Signal to Noise Ratio and Root Mean Square Error.