The Identification Algorithm of Forged Image Based on BP Neural Network

Author

Xing, Nan ; Zhu, Hong ; Ma, Wenqing ; Dong, Min

Author_Institution

Fac. of Autom. & Inf. Eng., Xi´´an Univ. of Technol., Xi´´an, China

fYear

2009

fDate

17-19 Oct. 2009

Firstpage

1

Lastpage

4

Abstract

In this paper, the characteristics of a class of pattern noise existing in a wide range of digital camera, which are called camera fingerprint, will be detected with a wavelet filter. Then the camera fingerprint models used to identify the authenticity of images is established by way of BP neural network. With this algorithm, it is not necessary to embed watermark into the images, and the results of identifying images that are forged with copy, paste and other common methods are effective.

Keywords

backpropagation; cameras; image processing; neural nets; watermarking; wavelet transforms; backpropagation; camera fingerprint; digital camera; forged image; identification algorithm; image authenticity; neural network; pattern noise; watermark; wavelet filter; wavelet transform; Digital cameras; Digital filters; Fingerprint recognition; Information filtering; Information filters; Neural networks; Watermarking; Wavelet coefficients; Wiener filter; Working environment noise;

fLanguage

English

Publisher

ieee

Conference_Titel

Image and Signal Processing, 2009. CISP '09. 2nd International Congress on

Conference_Location

Tianjin

Print_ISBN

978-1-4244-4129-7

Electronic_ISBN

978-1-4244-4131-0

Type

conf

DOI

10.1109/CISP.2009.5302690

Filename

5302690