Image re-compression detection method suitable for rail transit

Abstract

The invention discloses an image re-compression detection method suitable for rail transit, which includes modification of header file quantization table, JPEG image decompression, feature extraction, detector training and result analysis. Aiming at the problem of JPEG heavy compression in the existing rail transit image detection technology, the present invention intervenes in the JPEG image decompression process by artificially modifying the quantization table in the JPEG header file, reduces the occurrence of truncation errors, and designs corresponding post-processing The neural network improves the effectiveness of feature extraction, thereby improving the accuracy of the JPEG re-compression detection method, which is conducive to the rapid identification and detection of intelligent gates in rail transit.

Description

technical field [0001] The invention belongs to image processing technology, in particular to an image re-compression detection method suitable for rail transit. (Rail transit artificial intelligence security gate system) Background technique [0002] With the development of computer and network technology, the cost of image tampering has been greatly reduced. More and more digital images appear in daily life, leading to a decline in the credibility of digital images, so digital image forensics technology has attracted the attention of governments and law enforcement agencies. Currently, JPEG compression is a popular compression standard, and thus, forensic issues related to the JPEG format receive more attention. If part of the source image is moved into the destination JPEG image, it may cause JPEG recompression when saving again. There are already some methods for detecting JPEG recompression, and most JPEG forensics problems involve images saved in JPEG format. [00...

AI Technical Summary

Problems solved by technology

[0005] Purpose of the invention: the purpose of the present invention is to solve the deficiencies in the prior art, to provide a general image re-compression detection method suitable for rail transit, for the JPEG re-compression problem in the existing rail transit image detection technology, through Artificially modify the quantization table in the JPEG header file, intervene in the decompression process of the JPEG image, reduce the generation of truncation errors, and improve the effectiveness of feature extraction, thereby improving the accuracy of the JPEG re-compression detection method, which is beneficial to smart gates in rail transit Machine rapid identification and detection

Images