EFFICIENCY ANALYSIS OF TETROLET TRANSFORM-BASED FILTERING BY REMOVAL OF ADDITIVE NOISE IN IMAGES

Efficiency of filtering based on tetrolet transform for test image database with different properties distorted by additive white Gaussian noise with different intensity is investigated. As the performance criteria, both standard metrics, for instance, PSNR and visual quality metrics (PSNR-HVS-M, MSSIM, and FSIM) are used. Effect of test image features on optimal threshold is analyzed. A comparative analysis of the tetrolet transform-based filter with DCT-filter with respect toobject edge preservation and effective denoising is shown