An Improved Contrast Enhancement Algorithm for Infrared Images Based on Adaptive Double Plateaus Histogram Equalization

Shuo Li,Weiqi Jin,Li,Yiyang Li
DOI: https://doi.org/10.1016/j.infrared.2018.03.010
IF: 2.997
2018-01-01
Infrared Physics & Technology
Abstract:Infrared thermal images can reflect the thermal-radiation distribution of a particular scene. However, the contrast of the infrared images is usually low. Hence, it is generally necessary to enhance the contrast of infrared images in advance to facilitate subsequent recognition and analysis. Based on the adaptive double plateaus histogram equalization, this paper presents an improved contrast enhancement algorithm for infrared thermal images. In the proposed algorithm, the normalized coefficient of variation of the histogram, which characterizes the level of contrast enhancement, is introduced as feedback information to adjust the upper and lower plateau thresholds. The experiments on actual infrared images show that compared to the three typical contrast-enhancement algorithms, the proposed algorithm has better scene adaptability and yields better contrast-enhancement results for infrared images with more dark areas or a higher dynamic range. Hence, it has high application value in contrast enhancement, dynamic range compression, and digital detail enhancement for infrared thermal images. (C) 2018 Elsevier B.V. All rights reserved.
What problem does this paper attempt to address?