Share Email Print

Proceedings Paper

Locally adaptive contrast enhancement and dynamic range compression
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

In surveillance applications, the visibility of details within an image is necessary to ensure detection. However, bright spots in images can occupy most of the dynamic range of the sensor, causing lower energy details to appear dark and difficult to see. In addition, shadows from structures such as buildings or bridges obscure features within the image, further limiting contrast. Dynamic range compression and contrast enhancement algorithms can be used to improve the visibility of these low energy details. In this paper, we propose a locally adaptive contrast enhancement algorithm based on the multi-scale wavelet transform to compress the dynamic range of images as well as increase the visibility of details obscured by shadows. Using an edge detector as the mother wavelet, this algorithm operates by increasing the gain of low energy gradient magnitudes provided by the wavelet transform, while simultaneously decreasing the gain of higher energy gradient magnitudes. Limits on the amount of gain imposed are set locally to prevent the over-enhancement of noise. The results of using the proposed method on aerial images show that this method outperforms common methods in its ability to enhance small details while simultaneously preventing ringing artifacts and noise over-enhancement.

Paper Details

Date Published: 18 May 2012
PDF: 10 pages
Proc. SPIE 8355, Infrared Imaging Systems: Design, Analysis, Modeling, and Testing XXIII, 835518 (18 May 2012);
Show Author Affiliations
Robert Maschal, U.S. Army Research Lab. (United States)
S. Susan Young, U.S. Army Research Lab. (United States)

Published in SPIE Proceedings Vol. 8355:
Infrared Imaging Systems: Design, Analysis, Modeling, and Testing XXIII
Gerald C. Holst; Keith A. Krapels, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?