Research On Copy-Move Forgery Detection In Digital Images Based On Color Information

With the increasing popularity of digital image acquisition devices,digital images have become an integral part of human life.They are the main basis for news reports,court evidence,etc.,so the authenticity of images is very important.At the same time,with the emergence of a large number of digital image editing tools,the processing and modification of images has become more and more simple.To this end,effective technical means to identify the authenticity of digital images is necessary.This thesis introduces the principle of active detection technology and passive detection technology in the field of digital image forensics,and then focuses on the related technologies of copy-move forgery detection.For the digital image copy-move forgery,this thesis summarizes the characteristics and applicable scenarios of the existing detection algorithms,and then proposes two kinds of copy-move forgery detection algorithm based on the color information.Firstly,aiming at the problem of mismatching in color image copy-move forgery detection,this thesis proposes an image copy-move forgery detection method based on scale-invariant ORB(oFast and rBRIEF)algorithm and YIQ(NTSC luminance(Y)and chrominance(I and Q)color components)color feature.By extracting ORB features from Gauss scale space and adding YIQ color information features,ORB algorithm not only increases robustness to scale changes,but also effectively eliminates mismatching problems caused by the inconsistency of color information.The experimental results on CASIA digital image copy-move forgery detection database reveal that the algorithm has low mismatch rate and low computational complexity,and has strong robustness for post-processing operations such as blur,noise addition and JPEG compression.Secondly,aiming at the problem that the features extracted from the color image by existing copy-move forgery detection algorithms are not comprehensive and the matching time is too high,the blind identification algorithms for digital images using SURF(Speeded Up Robust Features)detector and quantized color features is studied.In the process of feature extraction,HSV quantization global color histogram and SURF detector are combined to form a hybrid feature that describes the content of color image comprehensively.K-Means and KNN method are used to improve matching efficiency in feature matching stage.The experimental results indicate that the algorithm effectivelydetects and locates the color image copy-move forgery on the CASIA 1.0 and FAU color image test libraries,and can detect multi-region tampering and multi-attack tampering of the image.The experimental result demonstrates that the proposed algorithm has higher detection accuracy and better matching time for color image copy-move forgery.Finally,the thesis summarizes the work of this thesis,and points out some possible research directions or key points in this field based on the existing research situation and the main limitations of current digital image forensics methods.