Distortion analysis of marked image and watermark in wavelet domain in AWGN channel with Information-Set Decoders and Gauss-Jordan transformation

This paper presents an analysis in the distortion level of the marked image and the watermark when it is submitted to different frequency sub-bands of the discrete wavelet transform 2D. This study also encodes the watermarked image with a code C(23,12), sends for AWGN channel and decodes with Information Set algorithm and Gauss-Jordan transformation, to verify the impact of this decoder on the distortion level of the images. Comparison is also made with a channel without encoder to visualize the Information Set decoder importance. Its function is to demonstrate the performance of the decoder such as have a smaller loss or distortion image as possible that the information will not endanger either watermarked image as watermark. Moreover, also report that depending of the frequency sub-band wavelet adopted, the images suffer different interference. The results demonstrate the high efficiency of Information Set decoder with Gauss-Jordan transformation, whose high PSNR values were got in the HH sub-band, justifying the choice of this sub-band for watermark inserting.