Authors: Dr. Devdas Saraswat, Pawan Singh Rajput
Abstract: Images are part of life that is a proof of many instants. Communication medium is quit vulnerable to transfer media as sensitive data may be perturb by intruders. So researcher work a lot for the network security optimization. This paper has proposed a model that authenticate the image obtained from the network. Work has proposed a model that extract optimize features from image by bio-inspired image segmentation. Segmented region of image was used for the embedding of authentic message in least significant bits region and used for training of mathematical model. Trained model is used for the prediction of clustered region when image undergoes in an attack. Experiment was done on real standard image dataset. Result shows that proposed model has improved the security under various attacks. Image quality was also highly maintained.
