Comparison of Performance Analysis of Image Enhancement in Neural Networks and Conventional Networks

Neural Networks have been developed rapidly during the recent few years and it is extensively applied for the enhancement of the digital image. An image enhancement is the process for improving the quality of digital image. A large number of conventional enhancement methods have been proposed and developed. The traditional methods are contrast enhancement, histogram equalization, edge sharpening, variety of filters etc. On the other hand, the back propagation algorithms with feed-forward networks, sigmoid functions, feed back networks are designed for enhancement in the digital image with high probability. The aim of this study is to reveal a comparison between conventional networks and Neural Networks. To accomplish this purpose, no of experiments has been conducted and examined. For this experimental analysis, statistical method has been used to classify and characterize the behavior of the images. Experiments on images are implemented to confirm the validity of the proposed analysis. One of the purpose of the study was to identify the main factor affecting the image and result are obtained were validated with existing techniques. This paper focuses on three popular features of image enhancement that are auto enhancement, face detection and edge detection.