Towards a new standard in medical video compression

Abstract

The objective of this article is to present a new video compression scheme with several potentialities in medical and industrial field. The proposed method will involve the process of transform, scaling and quantization, and it is based on bandelet transform coupled with the set partitioning in hierarchical trees (SPIHT) coding. The effectiveness of this technique results in combining two advantages. First it takes advantage of the anisotropic regularities of frames; second, it exploits the dependencies between the geometric transformed coefficients using SPIHT encoder to encode significant coefficients, while reducing redundancy, which, in turn, results in decreasing the volume of data without altering the pertinent information. The authors examined three scenarios, according to the above critical points of the visual quality of decompressed video. The first scenario concerns the choice of compression filter. The second identifies appropriate transform type, and the third validates the proposed algorithm with respect to classical algorithms. The performances of the proposed algorithm are evaluated using a set of objective video quality assessment metrics, including PSNR (peak signal- to-noise ratio), MSSIM (mean structural similarity) and VIF (visual information fidelity). The experimental results illustrate clearly the superiority of our algorithm with respect to Wavelet-SPIHT, MPEG-4, and H.264, H.265/MPEG-HEVC (video coding standards) when applied to a set of medical image tests.