Image coding using Periodic Walsh Piecewise-Linear transform

Abstract

The analysis of image coding using Periodic Walsh Piecewise-Linear (PWL) transform will be presented in this paper. The coding method employs an integer bit allocation scheme and Llyod-Max quantizers. Experimental results have shown that best results are obtained by modeling the de coefficient using a one-sided Gaussian distribution and the ac coefficients by using a Laplacian distribution. Comparison of coding real images using the periodic Walsh piecewise-linear transform and some orthogonal transforms will be given. Computer simulation results indicate that the image coding efficiency of the periodic Walsh piecewise-linear transform is close to that of Walsh and Haar transforms from the PSNR viewpoint.transforms commonly used, the Cosine transform has been shown to possess a superior energy compaction property [10].In this paper, we will present the results of coding real images using Periodic Walsh Piecewise-Linear PWL transform. Comparison with the some orthogonal transform coding is also considered. In section II, we will briefly present the Periodic Walsh Piecewise-Linear transform. Section III introduces the two-dimensional image transformation. In section IV, a description of the coding method will be given. Simulations and experimental results are introduced in section IV. Concluding remarks are drawn in section V.