Share Email Print

Journal of Electronic Imaging

Uneven illumination removal and image enhancement using empirical mode decomposition
Author(s): Soo-Chang Pei; Yu-Zhe Hsiao; Mary Tzeng; Feng Ju Chang
Format Member Price Non-Member Price
PDF $20.00 $25.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Uneven light distribution problems often arise in poorly scanned text or text-photo images and natural images taken by digital camera. An innovative image-processing technique for uneven illumination removal using empirical mode decomposition (EMD) is proposed. The EMD is local, adaptive, and useful for analyzing nonlinear and nonstationary signals. In this method, we decompose images by EMD and get the background level locally and adaptively. This algorithm can enhance the local reflectance in the image while removing uneven illumination for black/white text images, text-photo images, and natural color/gray-level images. The proposed technique can be very helpful for image and text recognition. The EMD can also be applied to the three color channels (RGB) of color images separately to estimate the reflectances of the three color channels. After we relight these channels using white light and the estimated reflectances, a simple color constancy task can be performed to correct certain poorly lighted color images. Our technique is compared with recently proposed methods for correcting images with uneven illumination and the experimental results demonstrated that the proposed approach can effectively enhance natural color/gray-level images and make text and text-photo images more readable under uneven illumination.

Paper Details

Date Published: 20 December 2013
PDF: 14 pages
J. Electron. Imag. 22(4) 043037 doi: 10.1117/1.JEI.22.4.043037
Published in: Journal of Electronic Imaging Volume 22, Issue 4
Show Author Affiliations
Soo-Chang Pei, National Taiwan Univ. (Taiwan)
Yu-Zhe Hsiao, National Taiwan Univ. (Taiwan)
Mary Tzeng, National Taiwan Univ. (Taiwan)
Feng Ju Chang, National Taiwan Univ. (Taiwan)

© SPIE. Terms of Use
Back to Top