Effective Detection Algorithm Of Recaptured Image

With today’s rapid development of Internet and wide application of big data and cloud computing technology, information resources have been in more prominent position, sharing of information resource becomes a necessary demand in the Internet age. As traditional information carrier, image plays the irreplaceable role. However, with rapid development of digital image acquisition equipment and processing software, we can easily get the false image, known as the distorted or the forged images. Today the content of image cannot be completely believed. In other words," seeing is believing" has gone far away, the spread of false information has seriously affected social trustworthiness in recent ten years. However digital image forensics technology has played a vital role. The technology is to analyze authenticity and integrity of image information. Based-LCD recaptured image detection belongs to the category of digital image forensics. The recaptured process can be divided into two steps, which firstly make original or distorted image shown on LCD screen, secondly capture the image shown on LCD screen to get the recaptured image with the same content as the original. This article first introduces the development background and theoretical framework of digital image forensics technology, then introduces significance and the process framework of recaptured image detection method, and finally conducts research from two aspects of imaging process interference and image quality with the help of related researches. We put forward two efficient detection algorithms based-LCD screen. In this paper, the main work is as follows.(1) The proposed recaptured image detection algorithm based on the process of image formation. Recaptured images are captured original or distorted image shown on the intermediaries. After shooting through two successive different scene, compression and noise can bring great interference to the recaptured image, the article use the image block-blur effect caused by camera JPEG compression mechanism and the image diffraction wave interference caused by recaptured equipments (LCD screen and camera) in the process of recapture. The experimental results show that the method can efficiently distinguish recaptured image and original image.(2) The proposed recaptured image detection algorithm based on image quality. The image quality inevitably falls after shooting through two successive different scene, we target character of image edge and region sharpness which are directly related to image quality to conduct research. We first use the non-sampling contourlet approach to "capture" singular characteristics of sub-band image edge to analyze statistical characteristic of the different directions sub-band; then analyze the character of region sharpness by estimating the matching degree parameters between image and the model based on the generalized gaussian distribution characteristics of natural images. The experimental results show that the method has very good effect for the recaptured detection.