Detection of copy–move image forgery using histogram of orientated gradients

ABSTRACT:

The increasing popularity of digital media and media editing software has led to widespread tampering of multimedia files for malicious purposes. The most common form of tampering associated with digital images is copy–move forgery, in which a portion of an image is duplicated and substituted in a different location. Thus, law enforcement and forensics experts require reliable and efficient means of detecting copy–move forgery. This paper proposes a blind forensics approach to the detection of copy–move forgery. The input image is segmented into overlapping blocks, whereupon a histogram of orientated gradients is applied to each block. Statistical features are extracted and reduced to facilitate the measurement of similarity. Finally, duplicated image blocks are detected by SVM. Experiment results demonstrate that the proposed method is able to detect multiple examples of copy–move forgery and precisely locate the duplicated regions, even when dealing with images distorted by translation involving small rotations, blurring, adjustment of brightness, and color reduction. We are currently working to improve detection in regions with rotation and scaling adjustment over large areas.

PROJECT OUTPUT VIDEO:

EXISTING SYSTEM:

• Most methods used in the detection of copy–move forgery can be categorized as either block-based methods or key point based methods.

• The first such method was proposed by Fridrich et al., using a block matching detection scheme based on discrete cosine transform (DCT).

• Key point-based methods differ from block-based methods in their reliance on the identification and selection of regions of high-entropy within an image (i.e. ‘‘keypoints’’). Some approaches involve the extraction of points of interest using a scale-invariant feature transform (SIFT), capable of detecting and describing clusters of points belonging to cloned areas. For example, Pan and Lyu estimated the transform between matched SIFT keypoints and searched all pixels within the duplicated regions after discounting the estimated transforms.

• Amerini et al. developed a SIFT-based method for the detection of copy–move attacks and transformation recovery.

• Costanzo et al. proposed three novel forensic detectors with the ability to remove global or local keypoints, based on anomalies in the distribution of keypoints following manipulation

DISADVANTAGES OF EXISTING SYSTEM:

• Most forgery detection methods are evaluated against simple forgeries that human viewers have no trouble identifying in low resolution images

• Lower accuracy.

PROPOSED SYSTEM:

The most important function of an algorithm for the detection of copy–move image forgery is determining whether a given image contains duplicated regions. Post-processing operations, such as rotation, scaling, blur degradation, and changes in contrast cannot be known a priori; therefore, it is computationally impossible to compare every possible pair of regions, pixel by pixel. The matching of overlapping blocks of a fixed-size could save time, particularly if augmented using feature extraction techniques such as DCT or PCA.

The proposed work having following steps,

1.The image is divided into overlapping blocks of a fixed-size.

2. Features of each block are extracted using HOG descriptors.

3. Similar block pairs are matched then SVM is used for find whether the image is forgery or real.

4. Finally bounding box is created for duplicated regions.

ADVANTAGES OF PROPOSED SYSTEM:

• The rate of accuracy is high ( 97.5 % ).

SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS:

• System          : Pentium Dual Core.

• Hard Disk      : 120 GB.

• Monitor          : 15’’ LED

• Input Devices : Keyboard, Mouse

• Ram                : 1GB.

SOFTWARE REQUIREMENTS:

• Operating system : Windows 7.

• Coding Language : MATLAB

• Tool                      : MATLAB R2013A

REFERENCE:

Jen-Chun Lee a, Chien-Ping Chang b, Wei-Kuei Chen, “Detection of copy–move image forgery using histogram of orientated gradients”, Information Sciences, ELSIEVER, 2017