Title :
Prediction of filtering efficiency for DCT-based image denoising
Author :
Abramov, S. ; Krivenko, S. ; Roenko, Alexey ; Lukin, Vladimir ; Djurovic, Igor ; Chobanu, Mikhail
Author_Institution :
Dept. of Signal Reception, Transm. & Process., Nat. Aerosp. Univ., Kharkov, Ukraine
Abstract :
The task of prediction practical efficiency of filtering on the basis of the discrete cosine transform (DCT) methods is considered. It is shown that it is possible to estimate the MSE values of images to be processed by means of calculation rather simple statistics of DCT coefficients. Moreover, the quasi-optimal value of threshold parameter for DCT filtering methods can be easy evaluated as well. The results are presented for different additive Gaussian noise levels and a set of gray-scale test images. Developed method provides an opportunity to decide is it worth applying filtering or not.
Keywords :
Gaussian noise; discrete cosine transforms; filtering theory; image denoising; mean square error methods; prediction theory; statistics; DCT coefficients; DCT filtering methods; DCT-based image denoising; MSE values; additive Gaussian noise levels; discrete cosine transform methods; filtering efficiency prediction; gray-scale test images; quasioptimal value; statistics; Boats; Estimation; Filtering; Streaming media; DCT; filtering efficiency; noisy image; prediction;
Conference_Titel :
Embedded Computing (MECO), 2013 2nd Mediterranean Conference on
Conference_Location :
Budva
DOI :
10.1109/MECO.2013.6601327
