Spiking neural network with blockchain for tampered image detection using forensic steganography images

Abstract

Accurate tools are required to acknowledge misleading images in order to maintain image legitimacy, and these tools must allow for legal operations on images. Additionally, after posting their images to the Internet, image owners lose rights over the images because there are no measures in place to safeguard them from misuse. One of the most well-liked techniques for addressing copyright disputes is the use of steganography technologies. The embedded steganography images can, sadly, be easily altered or deleted. To address this problem, this work presents the spiking neural network (SNN) with blockchain for tampered image detection utilizing forensic steganography images. Forensic steganography images that have been altered can be found with this SNN. Using steganography images from the database, SNN is trained in this model. The blockchain stores the owners’ access policies. The Python platform is used to implement the proposed strategy. F-measure, specificity, accuracy, precision, recall false positive rate (FPR), and false negative rate (FNR) are used to gauge how well the proposed approach performs. When compared to state-of-the-art approaches, the proposed approach obtained an impressive rise of 98.65%, in classification accuracy.