Welcome to the InfoSci Platform

Multiple Fusion Strategies in Localization of Local Deformation Tampering

Yongzhen Ke, Yiping Cui

Source Title: International Journal of Digital Crime and Forensics (IJDCF)13(2)

ISSN: 1941-6210|EISSN: 1941-6229|EISBN13: 9781799860358|DOI: 10.4018/IJDCF.2021030107

MLA

Ke, Yongzhen, and Yiping Cui. "Multiple Fusion Strategies in Localization of Local Deformation Tampering." IJDCF vol.13, no.2 2021: pp.103-114. http://doi.org/10.4018/IJDCF.2021030107

APA

Ke, Y. & Cui, Y. (2021). Multiple Fusion Strategies in Localization of Local Deformation Tampering. International Journal of Digital Crime and Forensics (IJDCF), 13(2), 103-114. http://doi.org/10.4018/IJDCF.2021030107

Chicago

Ke, Yongzhen, and Yiping Cui. "Multiple Fusion Strategies in Localization of Local Deformation Tampering," International Journal of Digital Crime and Forensics (IJDCF) 13, no.2: 103-114. http://doi.org/10.4018/IJDCF.2021030107

Export Reference

Favorite Full-Issue Download

View Full Text HTML

View Full Text PDF

Abstract

Tampering with images may involve the field of crime and also bring problems such as incorrect values to the public. Image local deformation is one of the most common image tampering methods, where the original texture features and the correlation between the pixels of an image are changed. Multiple fusion strategies based on first-order difference images and their texture feature is proposed to locate the tamper in local deformation image. Firstly, texture features using overlapping blocks on one color channel are extracted and fed into fuzzy c-means clustering method to generate a tamper probability map (TPM), and then several TPMs with different block sizes are fused in the first fusion. Secondly, different TPMs with different color channels and different texture features are respectively fused in the second and third fusion. The experimental results show that the proposed method can accurately detect the location of the local deformation of an image.