Research On Tampering Detection Of Seam Cropped Images

With the rapid development of the Internet,mobile phones,and computer,the image as a way of information dissemination,has been more efficient communication.At the same time,the development of technology also breeds some of the drawbacks.When the criminal use the software to delete the object in the image,it may lead to the distortion of the information transmitted by the image,misleading the public,and even endangering human health.Therefore,the detection of tampering images has become an urgent research.At present,there is an image processing technique called Seam-carving,which has been unanimously praised by removing or inserting a seam of the lowest energy to achieve mage scaling.Seam-carving is used for adjusting the size of image as well as realizing the removal of object in the image.Therefore,detecting the location of object deleted by seam-carving becomes a scientific challenge to the image forensics.According to the position movement law of the related pixel in the seam-carving process,the detection and location method of object deletion based on the BAG mispairing characteristic is proposed in this paper.Firstly,the BAG is extracted from the JPEG image;secondly,10-dimensional features of the BAG chart are extracted;thirdly,the classification result is obtained by using the non-supervision cluster method;finally,the location result is obtained according to the classification result chart.Experiment result shows that the proposed method can efficiently detect and locate the object deleted by seam-carving.In addition,for the location of the deleted object position,we proposed an automatic location algorithm to make up for the previous experimental results based on human visual.The method proposed in this paper is based on statistical computing.The three types of clustering results were vertically and horizontally superimposed and draw the corresponding histogram.Through a series of processing,we delete the position of the deleted object in the image,and mark it with the red box.The experimental results show that the method proposed in this paper can accurately locate the position of the deleted object and mark the corresponding position.