Method for detecting computer generated image and natural image based on wavelet transformation

Abstract

The invention relates to a method for detecting a computer generated image and a natural image based on wavelet transformation, which belongs to the technical field of image detection. The method comprises the following steps: firstly, performing color channel transformation to an image to be detected, transforming a RGB color image into an HSV color image; secondly, transforming all channels of the image further to obtain a multi-dimensional eigenvector including statistic matrix characteristics and high-frequency filter characteristics; thirdly, extracting characteristics by utilizing characteristics of the computer generated image and the natural images; fourthly, rapidly judging the truth of the image by a support vector machine (SVM), and detecting whether the image is the computer generated image or the natural image. The invention adopts a technology of combining the statistic matrix characteristics and the high-frequency filter characteristics, and has the characteristics of high precision, complete characteristics extraction, complete detection and the like, thereby greatly improving the detection precision.

Description

technical field [0001] The invention relates to a method in the technical field of image detection, in particular to a method for detecting computer-generated images and natural images using wavelet transform. Background technique [0002] The emergence of computer-generated images has become a serious obstacle to the credibility of current digital image forensics. Due to the characteristics of the human visual system or the limited business knowledge of image forensics technicians, highly simulated computer-generated images can easily be regarded as real digital images. images are used for image forensics. In the detection of computer-generated images, the most critical issue is how to accurately distinguish the feature differences between computer-generated images and natural images. Therefore, feature extraction is the primary problem in solving computer-generated images and natural image detection. The usual feature extraction uses a single type of feature information,...

AI Technical Summary

Problems solved by technology

[0003] Found through the retrieval of prior art document, Wen Chen, Yun Q.Shi and Guorong Xuan in document " Identifying computer graphics using HSV color model and statistical moments of characteristic functions " (" based on the computer image of HSV color model and characteristic function statistical moment Identification") (Multimedia and Expo, 2007 IEEE International Conference.USA: IEEE, 2007.1123-1126) (Multimedia Expo, 2007 IEEE International Conference), proposed a detection method based on the HSV color model, and used a classifier to judge the authenticity of the image, But the detection accuracy is not high, only 82.1%

And Ying Wang and Pierre Moulin in the document "OnDiscrimination Between Photorealistic and Photographic Images" ("Natural Image and Natural Image Identification") (Acoustics, Speech and Signal Processing, 2006.ICASSP.IEEE International Conference) (Acoustics, Language and Signal Processing Processing, 2006 ICASSP International Conference), proposed a new method for detection by obtaining two high-frequency features and one low-frequency feature by filtering. Although the detection accuracy has been improved, it can reach 90.2%, but it still cannot meet the actual application requirements.

Images