Title :
Digital Image Restoration Using Posterior Distribution and Updating Pixel by Self Threshold
Author :
Adi, Kusworo ; Mengko, Tati L R ; Suksmono, Andriyan Bayu ; Danudirdjo, Donny
Author_Institution :
Sch. of Electr. Eng. & Informatics, Inst. Teknologi Bandung
Abstract :
Metropolis Hastings Markov Chain Monte Carlo (MHMCMC) algorithm has been applied to simulate digital image restoration using posterior distribution and pixels update with self threshold method. This method was implemented to Lena image degraded with 0.15 th order salt and pepper noise by the variation of temperature at 1.5, 2.5 and 4.5. The result of digital image restoration in gray scale at temperature 1.5 is good image restored with DeltaSNR 13.072 dB. In this simulation, the number of Markov chains (1000 chains) and iteration (800 iteration) are fixed parameter
Keywords :
Markov processes; Monte Carlo methods; image restoration; image segmentation; statistical distributions; Lena image; Markov chains; Metropolis Hastings Markov Chain Monte Carlo algorithm; digital image restoration; fixed parameter iteration; pixels update; posterior distribution; self threshold; Degradation; Digital images; Image restoration; Low pass filters; Monte Carlo methods; Nonlinear filters; Pixel; Random variables; Stochastic resonance; Temperature; MHMCMC algorithm; Posteriori Distribution; Self Threshold;
Conference_Titel :
SICE-ICASE, 2006. International Joint Conference
Conference_Location :
Busan
Print_ISBN :
89-950038-4-7
Electronic_ISBN :
89-950038-5-5
DOI :
10.1109/SICE.2006.314638
