Image Forgery Detection Using MD5 and OpenCV

With the rise of digital media, image forgery has become increasingly prevalent, posing serious challenges in fields like journalism, forensics, and legal investigations. This project proposes a hybrid method for detecting image forgery by integrating MD5 hashing with advanced OpenCV-based image processing techniques. The approach begins by generating MD5 hash values to quickly verify file integrity; any mismatch between hashes of two images indicates potential tampering at the binary level.  For a deeper analysis, the system employs OpenCV to perform grayscale normalization, pixel-level difference mapping, and Structural Similarity Index (SSIM) analysis. These techniques help detect subtle changes in texture, luminance, and structure between the original and the suspect image. Adaptive thresholding and Gaussian blur are applied to enhance heatmaps, highlighting possible forged regions. The system further incorporates HSV color space analysis and frequency domain examination to uncover manipulations such as filtering or contrast adjustments. 

Additionally, a custom crop detection algorithm checks for changes in image dimensions to estimate crop percentage. The results are visualized through similarity metrics, forgery probability, and bar graphs. This user-friendly web application provides a reliable, automated solution for authenticating digital images, making it highly suitable for real-world applications.