Filtering based Image Decomposition and Restoration Approach

Abstract

In image processing, most of the time it is required to process the image by partitioning or decomposing it in different parts or representing it by mean of different features. Also, the quality of an acquired or received image is very much important from the further processing point of view. The partitioning or decomposition of the image and reconstruction of the original image from the distorted image are the prime areas of research when deals with the image filtering. Presented research work deals with the decomposition of the distorted color image and the restoration of the original color image. Average filtering is used for the decomposition of each grey level planes of the image in three components and later, the average and median filters are used to reconstruct the color image from these decomposed components of each grey level planes. Different experiments are carried out with the insertion of 0.01 to 0.05 variance Gaussian white noise (GWN). The proposed approach is evaluated on the basis of identified performance evaluation parameters, i.e., mean squared error; peak signal to noise ratio; signal to noise ratio; structural similarity index measure; and correlation coefficient. Presented image decomposition approach is lightweight from the implementation point of view and based on the obtained results, it is observed that the median filter produces the good result where small details are required in image restoration.