Due to the ease of image acquisition accessibility, the amount of visible contents shared in the form of image, video and text in daily newspapers, Whatsapp, Twitter, Facebook and Instagram is expanding exponentially nowadays. Recent advancements in image editing software make it simpler to alter images, thus increase to digital image forgery. In the past, efforts have been made by various researchers to determine the integrity and credibility of the image. Traditional approaches of image tempering detection may detect certain types of forgery by extracting simple features. On the other hand, Deep learning based techniques have transformed the field of digital image forgery detection and exhibited superior performance compared to traditional approaches due to their capability to automatically learn abstract features, which is particularly well-suited for detecting subtle and hidden manipulations in images. The fusion of traditional and deep learning techniques has significantly revolutionised digital image forgery detection, leading to more efficient, adaptable and accurate. The main objective of this research is to delve into the current techniques for detecting image forgery, the findings and analysis of several approaches, methodologies and performance achieved. Finally, it aims to encourage academicians to develop universal methods that can identify any form of manipulation in the given visual imagery.