Abstract: Images play an important in research and technology such as geographical information systems. Information is transmitted in the form of digital images and becoming a major method of communication in the modern age, but the image obtained is often corrupted with noise. Image processing plays a very important role in image denoising. Most of the applications make use of images in various fields such as medical, space exploring, aerial images, satellites images and many more. Images are received in defective conditions due to poor scanning and transmitting devices. Consequently, it creates problems for the process to read and understand such images. Before it can be used in applications the image that is received needs processing. To produce a visually high quality image, Image denoising is used which involves the manipulation of the image data. Selection of the denoising algorithm is dependent on application. Hence, the necessity to have detail information about the noise present in the image, to select the appropriate denoising algorithm is required. Image denoising issues can be addressed as an inverse problem. The paper presents an efficient denoising scheme by using LPG with PCA for better preservation of image local structure. LPG-PCA is compared by fast non- local means algorithm and the proposed method. In LPG-PCA, the pixel and its nearest neighbors are modeled as a vector variable. Trading samples are selected from the local search window by using block matching based on LPG. Such an LPG procedure guarantees that only the sample looks with similar content are used in the local statistics calculation. This calculations are for PCA transform estimation, so that the image local features can be well reserved. In order to accelerate the algorithm a Fast non- local means algorithm was developed to accelerate the calculation. Using proposed method the denoising performance is improved. Furthermore, results obtained by simulation using Matlab. We finally demonstrate the potential of the algorithms through comparisons. This paper specifies the description of noise, types of noise and LPG-PCA, Fast NLM and proposed algorithm that has been used for denoising the image.
Keywords: noise; types of noise; LPG-PCA algorithm; Fast NLM.