A Novel Approach to Enhancing Identity Document Authentication with Copy-Move Forgery Detection using CNN

The digital age facilitates easier transmission and sharing of official documents online, raising concerns about the integrity and confidentiality of sensitive information. This illicit activity involves unauthorized modifications or duplications of official documents for illegal use. Among these, image tampering is the most common activity for accessing unauthorized documents using copy-move forgery techniques. To provide integrity to personal documents, and to determine if a digital image is authentic or fake, the suggested schema created a Copy-Move Forgery Detection (CMFD) method utilizing a Convolution Neural Network deep learning technique. This paper introduces a CNN-based approach for CMFD for authenticating identity cards (ID) while transmitting over the internet. In the proposed model, a self-developed dataset (IDCCMFD-1000) is built to test the authenticity of the ID cards. The proposed model has been tested on various datasets, such as MICC-F2000 and IDCCMFD-1000, achieving a training accuracy of 98.33 %, 92 % and prediction accuracy of 96.54 % and 93.02 % respectively. These results underscore the model's effectiveness in distinguishing between authentic and forged images. Additionally, when the suggested model is contrasted with state-of-the-art schemes, the outcomes demonstrate better performance.