Discrete Laplace estimator with a variable moment order for the modified image reconstruction

In the paper, the discrete Laplace estimator with a variable moment order is considered. The application of this model in the image and video compression based on the discrete cosine transform (DCT) and quantization such as MPEG and JPEG standards is presented. Modeling the distribution based on the discrete variables collected at the decoder side can be used for the modified image reconstruction. The selection of the optimal moment order m for the discrete Laplace estimator with a variable moment order can outperform other discrete Laplace estimators in terms of PSNR criteria for different quality compression ratios and for the Y, U and V components.