Detection of Copy Move Forgery Using a Hybrid Algorithm

In recent years, image forgery detection in terms of copy-move detection has become a popular topic among researchers. For real-time images, this work presents a combination methodology of saliency detection and local binary pattern-based forgery detection. In a preliminary phase, saliency detection is employed to identify the forged portion. This forgery detection method can be applied in medical, forensics, and media to authenticate the integrity of the original image. It secures the modified portion's territory with the surrounding (global) area. The specific pixels/region for the tampered portions is then detected or captured by Local Binary Pattern features. This combined approach keeps the benefits of both the saliency map and the Local Binary Pattern, especially in terms of scaling or rotation, while also having a faster detection rate than the existing methods. In addition to the existing forgery detection result, the suggested method uses depth map information to properly identify the forged region. The severity of the tampered portion is assessed in terms of bits per pixel. The proposed method is unique in that it allows for the detection of tampered portions in real-time digital images. The proposed method is also shown to be considerably superior to existing methods in a state-of-the-art comparison.