Shape-Adaptive Transforms Filtering
Pointwise Shape-Adaptive DCT
EXPERIMENT RESULTS
All the results (images and tables) presented in this page have been obtained with our algorithms based on the low-complexity SA-DCT and can be replicated using the Pointwise Shape-Adaptive DCT Demobox.
For completeness, we also provide on a separate page some results which have been obtained by replacing the SA-DCT with the more complex “Shape-Adapted” DCT (obtained by orthogonalization of a set of DCT basis functions).
Denoising
Pointwise SA-DCT denoising
Results (PSNR, dB), noise is white additive Gaussian with variance σ².
(Click on the numbers to see the corresponding images)
Color Denoising
Pointwise SA-DCT denoising with structural constraint in luminance-chrominance space
Results (PSNR, dB), noise is white additive Gaussian with variance σ² on all RGB channels.
(Click on the numbers to see the corresponding images)
Deblocking
Pointwise SA-DCT deblocking and deringing for JPEG-compressed color images.
Results (PSNR, dB)
(Click on the numbers to see the corresponding images)
Deblurring
Pointwise SA-DCT regularized deconvolution
Results for four different deblurring experiments (see below)
(Click on the numbers to see the corresponding images)
Experiment 1 - Cameraman 256x256, blur: 9x9 uniform "box-car" blur, Gaussian white noise: BSNR 40dB (σ²=0.308)
Experiment 2 - Cameraman 256x256, blur: 15x15 1/(x²+y²), x,y=-7,...,7, kernel, Gaussian white noise: σ²=2
Experiment 3 - Cameraman 256x256, blur: 15x15 1/(x²+y²), x,y=-7,...,7, kernel, Gaussian white noise: σ²=8
Experiment 4 - Lena 512x512, blur: 5x5 separable [1, 4, 6, 4, 1]/16 filter, Gaussian white noise: σ²=49
back to top of page
TAMPERE UNIVERSITY OF TECHNOLOGY
-
Department of Signal Processing
