Image Forgery detection using

GLCNNB algorithms
 Abstract
• To develop a lightweight and fast technique for detecting image forgery.

• The algorithm will take less execution time for detecting region compare to
existing approaches.
• To introduce a GLCNNB which provide better result compare to other
reported algorithms if various attacks are applied and also able to detect
copy-move forgery region accurately with less execution time.
• To propose an efficient scheme for feature extraction from digital image,
these extracted features can be applied in many application areas of image
processing including image forgery detection.
 Introduction 
• We are living in an age where we are exposed to remarkable
array of visual imagery.
• Today's digital technology had begun to erode the integrity of
images.
• Over the past few years , the field of digital forensics has
emerged to restore some trust to digital images
 Image Forensic Tools
• Over the last few years, there has been a growing body of
work on tools for digital image forensics.
• These tools are capable of detecting tampering in images from
any camera, without relying on watermarks or specialized
hardware.
• Instead of watermarks, these tools assume that images possess
certain regularities that are disturbed by tampering.
 LITERATURE SURVEY
• Y. Zhao, S. Wang, X. Zhang, and H. Yao, “Robust hashing for image authentication using PCA
and local features,” IEEE Trans. Inf. Forensics Security, vol. 8, no. 1, pp. 55–63, Jan.2017.
Principal component analysis (PCA) is applied on fixed size image blocks (16 × 16, 32 × 32).
Further, eigenvalues as well as eigenvectors have been calculated of every block. Moreover,
lexicographical sort is applied and duplicated regions have been detected. This technique provides better
results compare to fridhrich methods. This technique is capable of detecting tampered region even if the
image is compressed as well as noisy.

• V. Christlein, C. Riess, J. Jordan, C. Riess, and E. Angelopoulou, “An evaluation of popular

copy-move forgery detection approaches,” IEEE Trans. Inf. Forensics Security, vol. 7, no. 6, pp.
1841–1854, Dec. 2018.
First, DWT is used to decompose image into four sub-bands. Further, low frequency sub band is
chosen for further processing. Moreover, singular value decomposition (SVD) is applied to extract feature
vector on reduced These vectors are matched and forged region detected. They have used gray scale as
well as color images for detecting duplicate region.
• Bianchi and A. Piva, “Image forgery localization via block-grained analysis of
JPEG artifacts,” IEEE Trans. Inf. ForensicsSecurity, vol. 7,no. 3, pp. 1003–
1017, Jun. 2018
Further, low frequency sub band is chosen and divided into non-overlapping blocks.
Moreover, the phase correlation utilize to calculate the spatial offset betwixt the copy-move
areas. Finally, pixel matching is employed for detecting tampered area. developed technique
provides better results even if the image is highly compressed and also considered as very
effective technique which is taking less computational time comparatively.

• H. Fu and X. Cao, “Forgery authentication in extreme wide-angle lens using

distortion cue and fake saliency map,” IEEE Trans. Inf.Forensics Security, vol.
7, no. 4, pp. 1301–1314, Aug. 2012.
First of all DCT coefficient have applied for extracting feature from image and then
Bayesian method is employed for locating possible forgery. Finally, the forgery region is
location successfully.
 EXISTING SYSTEM:

• They have found that the number of existing approaches, which are not able to detect
actual forgery region. So, the false matches are the major limitation. Detecting actual
region with good accuracy rate is the challenging tasks.

• We have processed the image block by block and extracted the feature vector for each
block then compared similar the feature vector to find the possible duplication.

• We have suggested a hybrid algorithm based on direct fuzzy transform and ring
projection which reduces the execution time as well as improve the accuracy.
 Proposed system
• To introduce a GLCNNB which provide better result compare to other reported
algorithms if various attacks are applied. And also able to detect copy-move forgery
region accurately with less execution time.
• To propose an efficient scheme for feature extraction from digital image, these
extracted features can be applied in many application areas of image processing
including image forgery detection.
 MODULES
• Input image

• Pre-processing

• Feature extraction

• Classification
BLOCK DIAGRAM
 Image Acquisition:

• Image acquisition in image processing can be broadly defined as the action of

retrieving an image from some source, usually a hardware-based source, so it can
be passed through whatever processes need to occur afterward.
• Performing image acquisition in image processing is always the first step in the
workflow sequence because, without an image, no processing is possible.
• The image that is acquired is completely unprocessed and is the result of whatever
hardware was used to generate it, which can be very important in some fields to
have a consistent baseline from which to work.
 Preprocessing and
segmentation
• For each and every process preprocessing is the necessary one, here the
preprocessing is done with the help of the artificial neural network.
• The artificial neural network consists of the three layers which are input layer,
hidden layer, output layer. In this each layer consists of the interconnected nodes, in
this each node will perform certain activation function
• The input image patterns are available to the network by means of the input layer.
This input layer communicates with the more number of hidden layers presented,
• The network where the actual processing is done, then these hidden layers unite to
form the output layer ,this output layer gives the output of the preprocessing system
 Feature Extraction:
• This is generally defined as the matrix, the features which are calculated by
using GLCM are mean square error, peak signal to noise ratio, entropy,
structural content, overall entropy.
 GLCMNN
• i. Create the framework matrix.

• ii. Compute the spatial relationship between the adjacent pixel.

• iii. Count the occurrences and fill the framework matrix.

• iv. Add the matrix to its transpose to make it symmetrical.

• v. Normalize the matrix.

 Classification:
• This Bayesian classifier is the simple probabilistic approach based on applying the

bays theorem. It has the advantage over the Bayesian MRF by recognizing the false

rate in high value.

• The classification process is done over the extracted features. The main novelty
here is the adoption of NB. NB classifier is applied over the features and the
classification is done.
• Naive Bayes classifiers are linear classifiers that are known for being simple yet
very .
• The probabilistic model of naive Bayes classifiers is based on Bayes’ theorem, and
the adjective naive comes from the assumption that the features in a dataset are
mutually independent.
• In practice, the independence assumption is often violated, but naive Bayes
classifiers still tend to perform very well under this unrealistic assumption.
Especially for small sample sizes, naive Bayes classifiers can outperform the more
powerful alternatives.
• Being relatively robust, easy to implement, fast, and accurate, naive Bayes
classifiers are used in many different ends. Some examples include the diagnosis of
diseases and making decisions about treatment processes , the classification of
RNA sequences in taxonomic studies , and spam in e-mail clients .
• However, strong violations of the independence assumptions and non-linear
classification problems can lead to very poor performances of naive Bayes
classifiers .
• We have to keep in mind that the type of data and the type problem to be solved
dictate which classification model we want to choose
• In practice, it is always recommended to compare different classification models
on the particular dataset and consider the prediction performances as well as
computational efficiency.
DATA FLOW
DIAGRAM:
USE CASE DIAGRAM:
 SOFTWARE
ENVIRNOMENT
• MATLAB is a high-performance language for technical computing. It integrates

computation, visualization, and programming in an easy-to-use environment where

problems and solutions are expressed in familiar mathematical notation.

• NEURAL NETWORK TOOLBOX:

• DATA ACQUISITION TOOLBOX:

output
 CONCLUSION
• In this project the forgery which occurs in an image has been
detected with the help of the GLCM.
• The GLCM which has to be processed on the image, those
images are already trained with the help of the neural network.
• Here the Bayesian classification is implemented along with the
GLCM to find the forged part of an image.
• In addition to that, certain characteristics features are evaluated
in this GLCM based forgery detection.
 REFERENCES
• S. Battiato, G. M. Farinella, G. M. Messina, and G. Puglisi, “Robust image alignment
for tampering detection,” IEEE Trans. Inf.Forensics Security, vol. 7, no. 4, pp. 1105–
1117, Aug. 2012.
• Y. Zhao, S. Wang, X. Zhang, and H. Yao, “Robust hashing for image authentication
using Zernike moments and local features,” IEEE Trans. Inf. Forensics Security, vol. 8,
no. 1, pp. 55–63, Jan.2013.
• I. Amerini, L. Ballan, R. Caldelli, A. Del Bimbo, and G. Serra, “A SIFT-based forensic
method for copy and move attack detection and Transformation recovery,” IEEE Trans.
Inf. Forensics Security, vol. 6,no. 3, pp. 1099–1110, Mar. 2011.
• P. Kakar and N. Sudha, “Exposing postprocessed copy-paste forgeries through
transform-invariant features,” IEEE Trans. Inf.Forensics Security, vol. 7, no. 3, pp.
1018–1028, Jun. 2012..
THANK YOU