Image tampering detection method based on Mask R-CNN

Abstract

The invention discloses an image tampering detection method based on an improved Mask R-CNN, and belongs to the technical field of image recognition. The method comprises the following steps: constructing an image tampering detection network based on the Mask R-CNN; the image tampering detection network comprises a main branch network, a noise branch network, a Resnet-FPN backbone network, a areaal proposal network RPN and a bilinear pooling ROI Align network. Inputting the tampered image into an image tampering detection network to perform feature combination on the input image classificationfeatures, noise features and tampering candidate area features, and outputting tampered image classification, tampered area positioning and image segmentation results; training and testing the imagetampering detection neural network by using the data set; and through the trained image tampering detection network, obtaining tampered image classification, tampered area positioning and image segmentation mask prediction. According to the invention, through an image tampering detection network based on Mask R-CNN, tampered images are classified, tampered areas are positioned, and manipulation areas are segmented, so that prediction of tampered image pixel levels is realized.

Description

Technical field [0001] The invention relates to the technical field of image recognition, in particular to a method for detecting image tampering based on Mask R-CNN. Background technique [0002] With the widespread adoption of high-resolution digital cameras and powerful digital image processing software, tampered pictures have become more real. As digital images are easily tampered with, a series of false image incidents have arisen. For example, when tamperers deliberately tamper with images for use in judicial forensics, news reports, and medical appraisals, the problems caused will cause immeasurable losses. Image stitching is one of the most common types of image forgery. It finds out two pixel points with marking characteristics, and uses corresponding technical means to gradually change characteristic pixels in one image into characteristic pixels in another image. [0003] The existing tamper detection methods can only infer whether a given image is forged, but cannot l...

AI Technical Summary

Problems solved by technology

[0003] Existing methods for tampering detection can only infer whether a given image is forged, but cannot locate stitched regions and segmentation mask regions at the same time

Images