An Efficient Approach for Identification of Copy- Move
Image Forgery using SURF and SIFT based Techniques
Shaik Rasool
1
, Uma N. Dulhare
2
, Vasavi Sravanthi Balusa
3
, Deepthi Joshi
4
and Shaziya jabeen
5
1, 3-5
Methodist College of Engineering and Technology, Hyderabad, India
Email: shaikrasool@outlook.com, vasavibvs@gmail.com, deepthijayam@gmail.com, shaziyajabeen@gmail.com
2
Muffakham Jah College of Engineering and Technology, Hyderabad, India,
Email: prof.umadulhare@gmail.com
Abstract—Internet has changed the world of computers and how we use them. Now there is no
field left that may not require internet. It has impacted mostly the fields like image processing,
machine learning, cyber security, data mining etc. Technological developments today have
made the life so comfortable and easy but also put forward a challenge of authentication of
digital data generated from various domains. This has become a major concern for security. To
address this concern, we are encapsulating facet point juxtaposition, adaptive over dissection
for forgery identification in our proposal. We have based our work on Key point and Section
based forgery identification techniques. Adaptive non-intersecting, irregular sections are
utilized to uncover suspicious sections in the images. Dissection algorithms are used to assist in
this process. Facet points are mined by comparing and juxtaposition each section with its facets.
Super pixels are used instead of facet points in the proposed forgery section mining system. The
neighbouring sections are melded into facet sections to acquire merged sections which have
similar color facets. To end the process morphological operations are used over merged sections
to obtain forgery section. The outcomes obtained shows that this proposed algorithm can
achieve superior and accurate outcomes even in most critical constraints when compared to
other existing approaches.
Index Terms— Image Processing, Forgery Identification, Image Forgery, Juxtaposition, Digital
Images.
I. INTRODUCTION
Digital image processing in this modern era has become one of the most protruding domains for research in
multidisciplinary fields. The core of electromagnetic spectrum and security is constituted by digital image
processing. It has become the primary focus of research with huge scope to safeguard privacy and maintain
confidentiality. Achieving the desired outcomes in image processing remains a challenging task that needs an apt
solution. Digital images are primary source for forensics, deep learning, AI and thrust areas. Using deep
learning techniques today, we can fake eminent personalities thus by posing challenge to forensics to detect
forgery images. The disparity between forgery and original image has been narrowed done deeply that it’s
almost becoming impractical to identify the difference. This drawback is addressed by proposed work [1].
Image forgery has become very common today in social networks. Copy-move forgery is among well-known
digital image manipulation practices today that copies portion of the image and is bonded at a different location
Grenze ID: 01.GIJET.9.1.703
© Grenze Scientific Society, 2023
Grenze International Journal of Engineering and Technology, Jan Issue