Image Forgery Detection Using MD5 & Open CV
Mohammad Shahnawaz Shaikh, Aparajita Biswal, Akruti Pandwal, Nilesh Khodifad, Bhavesh Atul Bhai Vaghela
Abstract
The more the advanced image making tools become accessible, the more image forgery is proliferated in hardware and software across all these domains – forensics, journalism, authentication, etc. In this research paper, we investigate methods on how to detect forged images by using Python programing, MD5 hashing algorithm and OpenCV library. The first part of the study utilizes MD5 to do an integrity verification by generating a unique hash value for each original image. The digital fingerprint you get from hashing each step of the image is this hash, and is a powerful way to compare against suspected forged images. MD5 is not perfect, but it is useful for performing an initial check of manipulated site; while not perfect, it does the trick. As powerful image processing software, OpenCV is used to enhance the detection process. This library provides detailed analysis of pixel patterns, possibly strange color discrepancies or just other things that would hint a manipulation. The proposed method performs advanced anomaly detection upon the extracted features from both original and suspect images with the aim of discovering inconsistencies meant to imply forgery. The research then proves through a series of experimental validations that the use of MD5 coupled with OpenCV significantly increases detection accuracy, while reducing false positives and detecting both overt and subtle manipulation. This not only underlines the need for creating a multilayered approach to counter the challenges posed by image forgery but also introduces a hybrid framework called UEFRG that combines three different methods done specifically to address image forgery. As digital content evolves, this study indicates the necessity of reliable detection methods to ensure integrity of visual media. The algorithm is further refined and machine learning techniques are explored for incorporation for future work to further improve the detection capabilities to keep pace with emerging threats in digital authenticity