الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 113 |  |  |
| | | from | 113 | | |
Abstract
Many types of noise can exist in images as a result of stratify different cedures on them like compression, data transmission and acquisition etc. servation of image specifics and attenuation of noise are the two important
Ijirements of image processing, but they are contradictory in nature. Most of the oising algorithms suggest in seemliness are either noise dependent or threshold emed. In real time ambience the type of noise in the image signal is unknown, i stratify an algorithm specific to noise, will never be successful under these nditions. So this research emphasis on the removal of noise while decreasing the 5S of specifics as low as possible and work effectively on any type of noise even high density.
Many filtering algorithm has low computational complication so that they can Iter noise in short time, and hence will find themselves proper for online and real me applications. The existing research works develop an efficient algorithm for amoving noise with high efficiency, yielding extremely low distortion in broad inge of noise densities with low run time and less computational complexity.
The traditional noise detection approaches used to apply the noise discovery irocess unconditionally on all the pixels without taking into consideration that the lixels are corrupted or not. The result is that the quality of image gets affected and he uncorrupted pixels still need to be filtrated out. The sol is to apply noise detection process before starting the filtration process so that only those pixels pass through the process of filtration which is targeted as uncorrupted pixels. The noise detection process presented in this thesis very simple working by moving the window pixel by pixel on all spatial resolution of image in the Binary Map, each value is observed. If it is either equal to the maximum intensity or equal to the minimum intensity as defined by the noise model then it is marked as noisy pixel (np) and indicated by ‘O’ in binary map. Also by applying the principle of ROLD.
• In this thesis, a novel work has been established an efficient algorithm to
remove noise has been established, inspired by the challenges that faced the other image denoising techniques.