Recognition Of Cloth Pattern Using SVM And Neural Network

Abstract

Digital image quality evaluation is the area of concern from last several decades and to evaluate the image quality there are two acclaimed approaches namely subjective quality evaluation and Objective quality evaluation approach which are recognized as standard approaches by International Telecom Union (ITU). The conventional reported works based on these acclaimed standard approaches are failed to achieve the accuracy. Existing approaches like structural similarity index matrix (SSIM) is an innovative approach which performs well on the content dependent distortions (content dependent in the sense variation in color, shape, texture) but not as efficient as Peak to Signal Noise ratio.
A novel algorithm is presented in this paper which provides an accurate performance to evaluate the quality of digital quality measurement. Mostly the conventional approaches fails to yield accurate results because most of the techniques relies on human visual data not on the residual distortion data which is termed as hidden distortion by most of the researchers. An innovative auto aggressive based on internal generative mechanism (IGM) which is successful to get Mean Square error both on distorted portion and residual portion as well. In order to yield the better result we have to combine the both residual distortion and visual distorted portion MSE’s. The experimental results yield the better performance as well as accuracy too.