Morphed Image Detection using Structural Similarity Index Measure

Abstract

Image authenticity has become a critical concern in the digital age, with the rise of image manipulation techniques. The proliferation of sophisticated image editing tools has made it increasingly challenging to distinguish between authentic and manipulated images. The proposed system employs image processing algorithms to analyze subtle distortions and alterations in images, aiming to identify morphed or Edited content. This project focuses on the detection of morphed or Edited areas in images using a combination of OpenCV and scikit-learn. The system utilizes OpenCV for image preprocessing and feature extraction, extracting relevant information from the images. Subsequently, scikit-learn is employed to develop a model capable of distinguishing between authentic and morphed images. The project aims to detect the morphed/Edited areas by extracting the features from the images and by comparing the images using SSIM algorithm. In addition to its standalone application, the system can be integrated into existing image verification pipelines, bolstering the overall security and trustworthiness of digital content.