This paper is published in Volume-2, Issue-4, 2016
Area
Digital Image Processing
Author
K. Sharath Chandra Reddy, Tarun Dalal
Org/Univ
CBS Group of Institution, Jhajjar, Haryana, India
Pub. Date
20 July, 2016
Paper ID
V2I4-1157
Publisher
Keywords
Forgery Detection Techniques, SVM Classifier, Hog Classifier and SASI Classifier

Citationsacebook

IEEE
K. Sharath Chandra Reddy, Tarun Dalal. Novel Approach for Image Forgery Detection Technique based on Colour Illumination using Machine Learning Approach, International Journal of Advance Research, Ideas and Innovations in Technology, www.IJARIIT.com.

APA
K. Sharath Chandra Reddy, Tarun Dalal (2016). Novel Approach for Image Forgery Detection Technique based on Colour Illumination using Machine Learning Approach. International Journal of Advance Research, Ideas and Innovations in Technology, 2(4) www.IJARIIT.com.

MLA
K. Sharath Chandra Reddy, Tarun Dalal. "Novel Approach for Image Forgery Detection Technique based on Colour Illumination using Machine Learning Approach." International Journal of Advance Research, Ideas and Innovations in Technology 2.4 (2016). www.IJARIIT.com.

Abstract

With the advancement of high resolution digital cameras and photo editing software featuring new and advanced features the chances of image forgery has increased. The images can now be altered and manipulated easily. Image trustworthiness is now more in demand. Images in courtrooms for evidence, images in newspapers and magazines, and digital images used by doctors are few cases that demands for images with no manipulation. Some forgery images that result from portions copied and moved within the same image to “cover-up” something are called as copy-move forgeries. In previous year author use different-different methods such as Principle Component Analysis (PCA), Discrete Wavelet Transform (DWT) & Singular Value Decomposition (SVD) are time consuming. In past many of the algorithm were failed many times in the detection of forged image. Because single feature extraction algorithm is not capable to contain the specific feature of the images. So to overcome the limitation of existing algorithm we will use meta-fusion technique of HOG and Sasi features classifier also to overcome the limitation of SVM classifier. Logistic regression would be able to classify the forged image more precisely.
Paper PDF