Signing Digital Images by ANN
Tarik Hajji · American Journal of Computational and Applied Mathematics · 2016
Signing digital images is a crucial operation in protecting the integrity of information in a digital image. In this work, we will show how we can use ANN as a agreeable hash function to the digital signature digital image. We will also give a description of an approach to develop a system of signing and verification for digital images using ANN. The objective of this work is not only to demonstrate the use of neural networks for signature but also to look for possible relationships between the rate of similarities images and to derive conditions of use for not collusion identical signatures. The advantages of this new approach are numerous; the shape of the standard signature is of fixed size for all documents, signature did not have kidney with formant data signed image and the signing operation is independent of any key signature. We will start with an introduction to the presentation of the state of the art on this topic. Then we will explain the methodology used to produce the signature system presenting the unsupervised learning algorithm and then we'll figurative results found, interpret and demonstrate this approach and finally we will conclude this study by all the conclusions and prospects for this work.