Unified Approach with Neural Network for Authentication, Security and Compression of Image: UNICAP

The present demands of scientific and social life forced image processing based applications tohave a tremendous growth. This growth at the same time has given number of challenges toresearcher to meet the desired objectives of either users or from solution perspectives. Amongthe various challenges, the most dominating areas are: reduction in required memory space forstorage or taken transmission time from one location to other, protection of image contents tomaintain the privacy and to facilitate the mechanism to identify the malicious modification if thereis any, either in storage or in transmission channel. Even though there are number of methodsproposed by various researchers and are existed as solutions, questions are remain open interms of quality, cost and complexity. In this paper we have proposed the concept based onneural network to achieve the quality of compression, protection and authentication all togetherusing the ability of universal approximation by learning, one way property and one to onemapping characteristics correspondingly. With the proposed methods, not only we canauthenticate the image but also positions of malicious activity given in the image can be locatedwith high precision. Proposed methods are very efficient in performance as well as carry thefeatures of simplicity and cost effectiveness.