VISVESVARAYA TECHNOLOGICAL UNIVERSITY

“JnanaSangama”, Belagavi – 590018

A SEMINAR REPORT
ON
“DIGITAL IMAGE FORGERY DETECTION”
Submitted in the partial fulfillment for the award of Bachelor of Engineering
degree in Computer Science & Engineering

Submitted By

SAHANA MP (4AD17CS078)

Under the Guidance of

Mrs. Nasreen Fathima

Associate Professor,
Dept. of Computer Science &Engg.
ATME College of Engineering,
Mysuru-570028

DEPARTMENT OF COMPUTER SCIENCE & ENGINEERING

ATME College of Engineering

13thKilometer, Mysuru-Kanakapura-Bengaluru Road,
Mysuru-570028

2022-2023
 VISVESVARAYA TECHNOLOGICAL UNIVERSITY
“JnanaSangama”, Belagavi – 590018

ATME College of Engineering

th
13 Kilometer, Mysuru-Kanakapura-Bengaluru Road, Mysuru-570028

Department of Computer Science and Engineering

CERTIFICATE

This is to certify that the Seminar Work entitled “Digital Image Forgery Detection” is the
bonafide work carried out by Sahana M P (4AD17CS078) in partial fulfillment for the award
of degree of Bachelor of Engineering in Computer Science and Engineering from
Visvesvaraya Technological University, Belagavi during the year 2022-2023.

Signature of the Guide Signature of the Coordinator Signature of the HoD

Mrs Nasreen Fathima Mrs Kavyashree E D Dr Puttegowda D
Associate Professor Assistant Professor Professor & HoD
Dept. of CSE Dept. of CSE Dept. of CSE
 ACKNOWLEDGEMENT

The successful completion of my work would be incomplete without the mention of

the names of the people who have made it possible. I am indebted to several individuals who
have helped me to complete the report.

I am thankful to Dr. L Basavaraj, Principal, ATME College of Engineering,

Mysuru, for having supported us in our academic endeavors.

I am extremely thankful to Dr. Puttegowda D, Associate Professor and HOD,

Department of Computer Science and Engineering, for his valuable support and his timely
inquiries into the progress of the work.

I express my earnest gratitude towards my seminar coordinator Mrs Kavyashree E D,

Assistant Professor, Department of Computer Science and Engineering, who helped me in
getting things done.

I am greatly indebted to my guide Mrs. NasreenFathima, Asst. Professor,

Department of Computer Science and Engineering, for her consistent co-operation and
support.

I am obliged to all teaching and non-teaching staff members of Department of

Computer Science and Engineering, Mysuru for the valuable information provided by them
in their respective fields. I am grateful for their co-operation during the period of our project.

Lastly, I thank almighty, parents and friends for their constant encouragement and
support, for helping me in completing the seminar successfully.

SAHANA M P (4AD17CS078)

i
 ABSTRACT

Digital image forgery detection examines the results of three image forgery detection
algorithms, Error Level Analysis, Copy-Paste Cloning Detection and Fourier based
Resampling Detection. Each algorithm was implemented within MATLAB and runon a
sample library of forged and unmodified images, including a selection of images from an
image manipulation dataset. Each method was found to have its own set of advantages and
limitations, however with all three methods combined.

The overall detection rate was an impressive 80%. Error Level Analysis provided decent
results on previously compressed, high-quality JPEG files, but struggled with newly
compressed images or low-quality samples. Copy-Paste Cloning Detection was highly
successful on images forged using cloning methods, however the overall runtime was
much higher than the other methods, and due to the nature of the algorithm false positives
were routinely detected.

Image Resampling Detection operated on a wide variety of images, provided good overall
results on each dataset, and the rate of false positives was low. The algorithm was also
highly efficient, however resampling must have occurred in order for any forgery to be
detected, and it was therefore unsuitable for direct copy-paste forgeries.

ii
 TABLE OF CONTENTS

Sl. No. Title Pg. No.

1. Acknowledgment i
2. Abstract ii
3. CHAPTER 1: INTRODUCTION
1.1: Background 7
1.2: Objective 8
4. CHAPTER 2: FEDERATED LEARNING
2.1: Applications of Digital Image Forensics 9
2.2: Types of image forgery detection 9
2.3: Active authentication 10
2.4 Passive authentication 11
Challenges of Digital Forgeries 12
5. CHAPTER 3: PASSIVE APPROACH
3.1: Copy-move forgery 13
3.2: Image-splicing 15
3.3: Image-retouching 16
6. CHAPTER 4: FORGERY DETECTION METHODS
4.1: Pixel-based image forgery detection 17
4.2: Format-based image forgery detection 17
4.3: Camera-based image forgery detection 18
4.4: Physical environment-based image forgery detection 19
4.5: Geometry-based image forgery detection 19
7. CONCLUSION 20
8. REFERENCES 21
 LIST OF FIGURES

Fig. No. Figure Name Pg. No.

2.1.1 Types of image forgery 9

2.3.2.1 Digital water marking 11

3.1.1 Example of copy-move forgery 13

3.1.2 Block diagram of copy-move forgery 14

3.2.1 Example of image-splicing 15

3.2.2 The common workflow of image splicing detection 16

3.3.1 Example of image-retouching 16

4.1.1 Pixel-based image forgery detection 17

4.2.1 Format-based image forgery detection 17

4.3.1 Camera-based image forgery detection 18

4.4.1 Physical environment-based image forgery detection 19

4.5.1 Geometry-based image forgery detection 19

 Digital Image Forgery Detection

CHAPTER 1

INTRODUCTION
Since the invention of photography, individuals and organisations have oftensought ways
to manipulate and modify images in order to deceive the viewer. Whilst originally a fairly
difficult task requiring many hours of work by a professional technician, with the advent of
digital photography it is now possible and fairly trivial for anyone to easily modify images,
and even easier to achieve professional looking results. This has resulted in wide reaching
social issues, ranging from the reliability of the images reported by the media to the
doctoring of photographs of models in order to improve their looks or body image. With
the sheer amount of methods available in which to manipulate an image, image forgery
detection has become a growing area of research in both academia and the professional
world alike.
Many methods exist in order to detect forgery within digital images, howeverit is difficult
to find which are the most efficient and practical to implement and run. Whilst one
algorithm may have a good detection rate, it could also have a large rate of false positives.
In addition, runtime is a major factor that contributes to the efficiency and overall usability
of an algorithm, but tends to only be mentioned academically as opposed to in real world
terms.

1.1 Background
In this day and age, digital images tampering has been made easy with widelyavailable
image editing softwares, such as Adobe Photoshop. The advancement of image editing
softwares has reached a level such that image tampering can be done without degrading its
quality or leaving obvious traces. This is alarming as imagesare now being presented as
supported evidences and historical records in various fields, such as in forensic
investigation, law enforcement, journalistic photography and medical images.
Moreover, in many instances tampered images have appeared in the news orsocial media,
such as the manipulated images of Iranian missile test released on the 9 thof July 2008 by
Sepah News, the official media arm of Iran’s Revolutionary Guard. The tampered image, is
aimed at exaggerating the country’s military capabilities. The tampered image made its
Dept. of CSE, ATMECE, MYSURU Page7
 Digital Image Forgery Detection

way into media circulation, making thefront page of notable newspapers, such as The
Financial Times and The Los Angeles Times. The forgery is detected a day later when the
same source released another image taken from the same angle at almost the same time, but
with different content.

1.2 Objective

The object of this project is to identify a Copy Move Forgery Detectiontechnique, with
both feature extraction and feature matching technique, capable ofobtaining better accuracy
rate while maintaining the computational time seen withSURF. Also, the proposed CMFD
method is compared with existing CMFD techniquesin terms of performance.

Dept. of CSE, ATMECE, MYSURU Page8

 Digital Image Forgery Detection

CHAPTER 2

DIGITAL IMAGE FORGERY

2.1 Applications of Digital Image Forensics

Digital forensics is commonly used in both criminal law and private investigation,
Forensic analysis the images on online social networks, Used for detecting tampered or
forged image, Image forgery detection system is needed in many fields for protecting
copyright and preventing forgery or alteration of images. It is applied in areas such as
journalism, scientific publications, digital forensic science, multimedia security,
surveillance systems etc.
The copy-move forgery is one of the difficult forgeries to detect in image processing. It is
common image tampering technique used now a day. In this some part of the image needs
to be covered to add or remove information of an image.

2.2 Types of image forgery detection

For authentication of images several methods have been developed. These methods are
broadly categorized into two classes: Active authentication and Passive authentication.

Fig 2.1.1 Types of Image Forgery

Dept. of CSE, ATMECE, MYSURU Page 9

 Digital Image Forgery Detection

2.3 ACTIVE AUTHENTICATION

In active authentication techniques prior information about the image is
indispensable to the process of authentication. It is concerned with data hiding where
some code is embedded into the image at the time of generation.
Verifying this code authenticates the originality of image. Active authentication methods
are further classified into two types digital watermarking and digital signatures. Digital
water marks are embedded into the images at the time of image acquisition or in
processing stage and digital signatures embed some secondary information, usually
extracted from image, at the acquisition end into the image. The main drawback of these
approaches remains that they are to be inserted into the images at the time of recording
using special equipment’s thus prior information about image becomes indispensable.
There are two types:
2.3.1 Digital signature
Digital signature is one among the active method used for detecting image forgery or
tampering. Digital signatures embed some secondary information, usually extracted from
image, at the acquisition end into the image.
This one is mathematical approachapplied to confirm the integrity and authenticity.
Digital signature comes under active methods applied for detection of image
forgery.Digital Signature equivalent to the handwritten signature in which itpossess a key
or signature, a digital signature provideexcessimplicit security.A secret key X is applied
to produceYarbitrary matrices with entries homogeneously divide up in the range [0, 1]
in Digital Signature.In authors proposed a method of a low phase filter usedforevery
random matrix recurrently to acquired X random pattern.
The model generates a digital signature using through the signing operation to the image.
Imagesigningoperation contains the under-mentioned phase:-
i) By applying parameterized wavelet feature images are decomposed.
ii) Extract the Standard Digital Signature.
iii) The crypto signature producesthrough the private key and hash the extracted the
Standard Digital Signature Cryptographically.
iv) Digital images along with crypto signature are delivered to the client.

Dept. of CSE, ATMECE, MYSURU Page 10

 Digital Image Forgery Detection

2.3.2 Digital Watermarking

Digital water marks are embedded into the images at the time of image acquisition or in
processing stage.Digital watermarking is a technology in which identification
information is embedded into the data carrier in ways that cannot be easily noticed, and
in which the data usage will not be affected. This technology often protects copyright of
multimedia data, and protects databases and text files.

Fig 2.3.2.1 Digital Watermarking

2.4 PASSIVE AUTHENTICATION

Passive authentication also called image forensics is the process of authenticating images
with no requirement of prior information just the image itself. Passive techniques are based
on the assumption that even though tampering may not leave any visual trace but they are
likely to alter the underlying statistics. It is these inconsistencies that are used to detect the
tampering. Passive techniques are further classified as forgery dependent methods and
forgery independent methods.
Forgery dependent detection methods are designed to detect only certain type of forgeries
such as copy-move and splicing which are dependent on the type of forgery carried out on
the image while as forgery independent methods detect forgeries independent of forgery
type but based on artifact traces left during process of re-sampling & due to lighting
inconsistencies. The main objective of passive detection technique remains to classify a
given image as original or tampered. Most of the existing techniques extract features from
image after that select a suitable classifier and then classify the features.

Dept. of CSE, ATMECE, MYSURU Page 11

 Digital Image Forgery Detection

2.5 Challenges of Digital Forgeries

Some of the major challenges are

1. Data Provenance.

2. Migration of digital information.

3. Ethical, legal and institutional issues.

4. Differentiating Forgery and clarity modification.

5. Benchmarking and Standard data set.

Dept. of CSE, ATMECE, MYSURU Page 12

 Digital Image Forgery Detection

CHAPTER 3

PASSIVE APPROACH
3.1 Copy-Move Forgery
Copy-move is one of the most widespread image tampering technique, also it is very
difficult to identify this type forgery as the copied image is taken from the same image. “In
Copy-Move image forgery, a part of the image is copied and pasted to another part of the
same image. It simply requires the pasting of image blocks in same image and hiding
important information or object from the image.” This method involves copying of some
area of an image superimposing it on some other area of the same image.

Fig 3.1.1 Example Of Copy-Move Forgery

As the area copied belongs to the same image, the dynamic range and color remains will be
same as the other part of image. Copy-move forgery is usually done to either hide some
part of the image of to show some part of the image multiple times. Copy-move forgery
detection aims at detecting the same or extremely similar areas and a lot of methods have
been proposed to solve this problem. Generally, these detection methods are summarized
into two categories: block matching-based and point matching-based. When creating a
Copy-move forgery, it is often necessary to add or remove important features from an
image. To carry out such forensic analysis, various technological instruments have
Dept. of CSE, ATMECE, MYSURU Page 13
 Digital Image Forgery Detection

developed.

Dept. of CSE, ATMECE, MYSURU Page 23

 Digital Image Forgery Detection

Fig 3.1.2 Block Diagram of Copy-Move Forgery

 Framework of the Proposed Algorithm.

The discussion above draws forth the framework of CMFD that is described. The steps of
the proposed CMFD technique are given as follows:
(1) Dividing the grayscale image into fixed sized overlapping blocks.
(2)Applying DCT toeach extracted block.
(3) Extracting Gaussian RBF kernel PCA-based features from each DCT square block.
(4) Matching similar block pairs.
(5) Removing the isolated block and output the duplicated regions.
For the implementationof proposed method, the algorithm is applied over the grayscale
images.

Dept. of CSE, ATMECE, MYSURU Page 14

 Digital Image Forgery Detection

3.2 Image Splicing

Image splicing is a technology of image compositing by combining image fragments from
the same or different images without further post-processing such as smoothing of
boundaries among different fragments. Image splicing forgery involves composition or
merging of two or more images changing the original image significantly to produce a
forged image.
In case images with differing background are merged then it becomes very difficult to
make the borders and boundaries indiscernible. Splicing detection is a difficult problem
whereby the composite regions are investigated by a variety of methods. Abrupt changes
between different areas that are combined and their backgrounds provide valuable traces to
detect splicing in the image under consideration.

Fig 3.2.1 Example Of Image Splicing Forgery

In order to get a large field of view image, we need to establish visual correlation among
the images which were acquired by cameras placed at different locations. Via matching
principle, we should com-pare the level of similarity between the target area and the same
size area from the different search area in images, then we need to identify the position,
where is the highest level of similarity. The position viewed as the best splicing
position. Selecting two images from the images captured by cameras, then two sets of
pixels were selected at intervals of a certain distance in horizontal direction.

Dept. of CSE, ATMECE, MYSURU Page 15

 Digital Image Forgery Detection

Fig 3.2.2 The common workflow of image splicing detection.

Image splicing forgery involves composition or merging of two or more images changing
the original image significantly to produce a forged image. In case images with differing
background are merged then it becomes very difficult to make the borders and boundaries
indiscernible. Splicing detection is a difficult problem whereby the composite regions are
investigated by a variety of methods. Abrupt changes between different areas that are
combined and their backgrounds provide valuable traces to detect splicing in the image
under consideration.

3.3 Image Retouching

In Image Retouching, the images are less modified. It just enhances some features of the
image. There are several subtypes of digital image retouching, mainly technical retouching
and creative retouching.

Fig 3.3.1 Example Of Image Retouching Forgery

Dept. of CSE, ATMECE, MYSURU Page 16

 Digital Image Forgery Detection

CHAPTER 4

FORGERY DETECTION METHODS

.

4.1 Pixel-based image forgery detection

Pixel-basedtechniques emphasize on the pixels of the digital image. These techniques are
roughly categorized into four types. We are focusing only two types of techniques copy-
move and splicing in this paper. This is one of the most common forgery detection
techniques.

Fig 4.1.1 Pixel-based image forgery detection

4.2 Format-based image forgery detection

Formatbasedtechniques are another type of image forgery detection techniques. These are
based on image formats and work mainly in the JPEG format. These techniques can be
divided into three types. If the image is compressed then it is verydifficult to detect forgery
but these techniques candetect forgery in the compressed image.

Fig 4.2.1 Format-based image forgery detection

Dept. of CSE, ATMECE, MYSURU Page 16

 Digital Image Forgery Detection

4.3Camera-based image forgery detection

Whenever we capture an image from a digital camera, the image moves from the camera
sensor to the memory and it undergoes a series of processing steps, including quantization,
colour correlation, gamma correction, white balancing, filtering, and JPEG compression.
These processing steps from capturing to saving the image in the memory may vary on the
basis of camera model and camera artifacts.These techniques work on this principle. These
techniques can be divided into four categories as shown

Fig 4.3.1 Camera-based image forgery detection

4.4Physical environment-based image forgery detection

Consider the creation of a forgery showing two movie stars, rumoured to be romantically
involved, walking down a sunset beach. Such an image might be created by splicing
together individual images of each movie star. In so doing, it is often difficult to exactly
match the lighting effects under which each person was originally photographed.
Differences in lighting across an image can then be used as evidence of tampering. These
techniques work on the basis of the lighting environment under which an object or image is
captured.Lighting is very important for capturing animage. These technique are divided
into three categoriesas shown.

Dept. of CSE, ATMECE, MYSURU Page 16

 Digital Image Forgery Detection

Fig 4.4.1 Physical environment-based image forgery detection

4.5Geometry-based image forgery detection

Grooves made in gun barrels impart a spin onto the projectile for increased accuracy and
range. These grooves introduce somewhat distinct markings to the bullet fired, and can
therefore be used to link a bullet with a specific handgun. In the same spirit, several image
forensic techniques have been developed that specifically model artifacts introduced by
various stages of the imaging process. Geometry-based techniques make measurement of
objects in the world and their position relative to the camera. Geometry-based image
forgery techniques are divided into two categories.

Fig 4.5.1 Geometry-based image forgery detection

Dept. of CSE, ATMECE, MYSURU Page 19

 Digital Image Forgery Detection

CONCLUSION
Image forgery is a treat big threat as new and new tools are available with cheaper price for
forging digital image. As there are many types of image forgeries, viz, copy-move forgery,
image splicing, image retouching, and lighting condition, it is very difficult to have a
image forgery identification techniques which applies to all types of forgeries. There are
few forgeries like copy-move forgery, image retouching and image splicing, which are
intentionally done with malafide intentions while use of lightning conditions is mostly
done for enhancing image or to remove noise came due to bad image source of error while
image capturing.
Increasingly significance of credibility of digital resources is becoming questionable? The
need for assurance of authentic digital resources is more vital than ever. Ultimately it is
hoped that due to increasing importance new significant problems of research and its
associated challenges for researchers and practitioners in the digital communities soon
become major area for research.

Dept. of CSE, ATMECE, MYSURU Page 20

 Digital Image Forgery Detection

REFERENCES
Journals

[1] AmanpreetKaur and Richa Sharma, "Optimization of Copy-Move Forgery Detection

Technique," International Journal of Advanced Research in Computer Science and
Software Engineering, vol. 3, no. 4, April 2013.
[2] Zhen Zhang, Ying Zhou, Jiquan Kang, and Yuan Ren, "Study of Image Splicing
Detection," Advanced Intelligent Computing Theories and Applications. With Aspects of
Theoretical and Methodological Issues, vol. 5226, pp. 1103-1110, 2008.
[3] Susama G Rasse, "Review of Detection of Digital Image splicing Forgeries with
illumination color estimation," International Journal of Emerging Research in
Management & Technology, vol. 3, no. 3.
[4] P. SabeenaBurvin and J. Monica Esther, "Analysis of Digital Image Splicing
Detection," IOSR Journal of Computer Engineering, vol. 16, no. 2, pp. 10-13, Mar-Apr
2014.
[5] Micah K. Johnson and HanyFarid, "Exposing digital forgeries by detecting
inconsistencies in lighting," ACM Multimedia and Security Workshop, 2005.

Web References
[1]. Google Pubs. - https://research.google/pubs/pub45808/ [Accessed on: 30/3/2020]
[2]. Google Blogs - https://ai.googleblog.com/2017/04/federated-learning
collaborative.html- [Accessed on: 24/3/2020]

[NOTE : ADD MINIMUM 20 REFERENCES]

You can refer your IEEE base paper references.

Dept. of CSE, ATMECE, MYSURU Page 21