DETECTION OF FAKE COLORIZED IMAGES USING DENSE CONVOLUTION  NETWORK

S. Aashmi Miranda; Dr. K. Thulasimani

DETECTION OF FAKE COLORIZED IMAGES USING DENSE CONVOLUTION NETWORK

Authors: S. Aashmi Miranda , Dr. K. Thulasimani

Unique Paper ID: 150218
Volume: 7
Issue: 4
PageNo: 104-110

Keywords: image forgery image detection histogram feature encoding neural network.

Abstract:
- Image forgery implies altering the digital image to some meaningful or valuable data. Image forensics is a well developed field that analyzes the images of specific conditions to build up trust and genuineness. Although image editing techniques can provide signiï¬cant appreciation of the image or entertainment value, they may also be used with malicious intent. An emerging image editing technique is colorization, in which gray scale images are colorized with realistic colors. But this technique may also be intentionally applied to certain images to confound object recognition algorithms. In this work, it is observed that, colorized images, usually change the images using a variety of mechanisms. The digital image developed from the colorization Method possess statistical differences in their RGB channels, hue and saturation channels and also need to observe statistical inconsistencies in the dark and bright channels, because the colorization process will mainly affect the dark and bright channel values. Based on the experiments in the hue, saturation, dark and bright channels, two simple yet effective detection method Histogram based and Feature Encoding based Fake Colorized Image Detection along with the dense convolution network are proposed for detecting the fake colorized images

Download article

email to a friend

Cite This Article

ISSN: 2349-6002
Volume: 7
Issue: 4
PageNo: 104-110

DETECTION OF FAKE COLORIZED IMAGES USING DENSE CONVOLUTION NETWORK

Available:https://ijirt.org/Article?manuscript=150218

Impact Factor
8.01 (Year 2024)

UGC Approved
Journal no 47859

Join Our IPN

IJIRT Partner Network

Submit your research paper and those of your network (friends, colleagues, or peers) through your IPN account, and receive 800 INR for each paper that gets published.

Join Now