Illuminant colour based image forensics using mismatch in human skin highlights

Author

Francis, Kiran ; Gholap, Sandeep ; Bora, Prabin Kumar

Author_Institution

Dept. of Electron. & Electr. Eng., Indian Inst. of Technol., Guwahati, Guwahati, India

fYear

2014

fDate

Feb. 28 2014-March 2 2014

Firstpage

1

Lastpage

6

Abstract

This paper presents a method to detect image splicing, specifically copying and pasting a person from an image to another. This is based on the fact that it is very difficult to match the illuminance property of the light sources estimated from several persons in a spliced image. The proposed method detects the illuminant colour mismatch among different persons in a composite image. The dichromatic reflection model is utilized to obtain the illuminant colour estimate from human skin highlights. The illuminant colour obtained is quantified using chromaticity coordinates. It is then matched against that of different persons in the composite image to detect the forgery.

Keywords

image colour analysis; image forensics; image retrieval; lighting; skin; chromaticity coordinates; colour estimate; composite image; dichromatic reflection model; human skin highlights; illuminant colour mismatch; image forensics; image splicing; Color; Eigenvalues and eigenfunctions; Image color analysis; Nose; Principal component analysis; Skin; Vectors; Dichromatic reflection model; Illuminant mismatch; Image forensics;

fLanguage

English

Publisher

ieee

Conference_Titel

Communications (NCC), 2014 Twentieth National Conference on

Conference_Location

Kanpur

Type

conf

DOI

10.1109/NCC.2014.6811349

Filename

6811349