Multiple Fusion Strategies in Localization of Local Deformation Tampering

Multiple Fusion Strategies in Localization of Local Deformation Tampering

Yongzhen Ke, Yiping Cui
Copyright: © 2021 |Pages: 12
DOI: 10.4018/IJDCF.2021030107
Article PDF Download
Open access articles are freely available for download


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.
Article Preview


The rapid development of image processing software makes it easy for people to use the retouching software to tamper with the image according to their wishes. Tampering images may be used by criminals to publish false information, and widespread dissemination of false information may cause social panic. Tampering with images can also involve areas of crime such as public opinion and violations of portrait rights. The purpose of news work is to publicize the truth, not to mention the use of tampered images to attract people's attention. For entertainment or beauty, people will modify photos before posting them on social networks. Many magazines and businesses also use unrealistic images to attract consumers. For example, many photos of the stars we saw in magazines have been modified with retouching software. This modified beauty will attract the public's attention and cause the public's pursuit of this beauty, but this beauty may not be in line with the health of the human body. Excessive pursuit of this unhealthy beauty will affect people's physical and mental health. In addition, ads such as fitness and weight loss often use false information to attract consumers to buy their products, which is unfair to consumers. People have long-term seen this over-modified "beautiful" images, which makes person mistakenly believe that this modified figure is "normal". This misunderstanding may affect people's values, and at the same time, the pursuit of such excessive "beauty" will affect people's physical and mental health. Therefore, it is necessary to forensic the authenticity of the image.

Image local deformation is one of the common methods of image tampering, including three basic operations of local translation, local scaling, and local rotation (Andreas Gustafsson, 1993). Through the use of Liquefy Tools like Photoshop, people can freely distort any area of the image and change the shape of objects in the image. People also use this tool to make the people whom in the figure look thinner and so on, an example is shown in Figure 1. Figure 1 shows that Forward Warp Tool in the Liquefy Tool is used to achieve translational deformation, Figure 1(a) is the original image without deformation, Figure 1(b) shows the process of using the Forward Warp Tool, and Figure 1(c) shows the tampered image of deformation. It can be seen that the arms of the woman in the image are deformed. The liquefaction tool tampers the image through a circular selection, which makes the pixel values in the middle of the selected area a great change while the pixel values at the edge of the selected area almost unchanged. Multiple basic deformations will be applied in practical applications.

Figure 1.

An example of local deformation


Complete Article List

Search this Journal:
Volume 15: 1 Issue (2023)
Volume 14: 3 Issues (2022)
Volume 13: 6 Issues (2021)
Volume 12: 4 Issues (2020)
Volume 11: 4 Issues (2019)
Volume 10: 4 Issues (2018)
Volume 9: 4 Issues (2017)
Volume 8: 4 Issues (2016)
Volume 7: 4 Issues (2015)
Volume 6: 4 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing