Passive contrast enhancement detection using NSCT-based statistical features and ensemble classifier
by Gajanan K. Birajdar; Vijay H. Mankar
International Journal of Electronic Security and Digital Forensics (IJESDF), Vol. 14, No. 4, 2022

Abstract: Due to widespread use of digital images and sophisticated image editing software, it is quite easy to create digital image forgeries without leaving any visual traces of doctoring. Contrast enhancement (CE) processing is popularly used to hide the traces of doctoring in copy-and-move image forgery operation by malicious users. In this paper, global blind contrast enhancement detection algorithm is proposed using various statistical parameters based on Gaussian distribution and generalised Gaussian distribution features, energy and grey level run length matrix (GLRLM) descriptors after NSCT decomposition. Fisher feature selection criterion is utilised to choose the most relevant features and to remove the less important features. Detection accuracy of the algorithm is investigated using various ensemble classifiers architectures. Experimental results are presented using four different ensemble classifier architectures class-I to class-IV for C_b and grey image database. The proposed algorithm outperforms all the existing feature-based approaches compared using the detection accuracy.

Online publication date: Mon, 04-Jul-2022

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Electronic Security and Digital Forensics (IJESDF):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com