Design Of Image Forensics System Based On Embedded Platform

Digital images,as one of the most important carrier for the acquisition and dissemination of information during the rapid development of information technology,have been widely used in our daily life.As well as the booming of the digital image processing technology,the digital image processing and tampering softwares are being more and more powerful.These softwares bring convenience and enjoyment to people.However,if someone does the malicious tampering to change the image information on purpose,it will result in the negative influence on the justice of judicial,economic,social,academic and other fields to a great extent.Therefore,the image forensics technology has extensive value of application and far-reaching research significance.So far,there are a lot of state-of-the-art researches and techniques of digital image forensics.But few of them aiming the application of image forensics for widespread usage,especially the application of embedded devices.With low cost,low power consumption and easy to carry,the application of image forensics on embedded devices is of great beneficial to the widespread used for image forensics.This thesis combined with the development technology of embedded side,study and realized the image forensics system based on embedded platform.The main contribution of this thesis includes:1)We selected two kinds of single JPEG compression detection and double JPEG compression detection algorithms with good performance and low computational complexity(it is suitable for realization at the embedded side).Then,we combined with block and sliding window method with C++ programming,which achieved a more accurate location of the image tampering area.According to the development technology of ARM embedded side and the designing of the detection program,to implement an image forensics software based on single JPEG detection on ARM embedded side.2)In order to improve the shortages of the image forensics based on ARM side,such as hard to manipulate,poor UI,etc.We choose Android platform to realize the forensics system.At first,we construct the image tampering detection framework based on NDK,Open CV and SVM on Android side.In this framework,the module of single JPEG compression detection and double JPEG compression detection were implemented.Based on these two detection modules enable to identify the authenticity of non-JPEG and JPEG images,respectively.And finally,it can be locate the image area where maybe tampered.This thesis designed and implemented the system of image tampering detection and localization on ARM and Android platform,respectively.It is verified that the image forensics system based on the embedded platform can effectively judge the authenticity and originalness of image through systematically testing.And the performance of the forensics system is able to meet the practical requirement.