Noniterative Power-Constrained Contrast Enhancement Algorithm for OLED Display
SCIE
SCOPUS
- Title
- Noniterative Power-Constrained Contrast Enhancement Algorithm for OLED Display
- Authors
- Chan Young Jang; Suk-Ju Kang; Kim, YH
- Date Issued
- 2016-11
- Publisher
- IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
- Abstract
- This paper proposes a new power-constrained contrast enhancement algorithm for organic light-emitting diode displays that provides high perceptual image quality with low power consumption. Unlike existing methods, the proposed method can achieve the target power consumption (TPC) without iteration, while significantly alleviating overenhancement artifacts using global and local attributes compared to the benchmark methods. The proposed method consists of three steps. In the first step, mapping functions, which map the input intensity into the output intensity, are generated considering the TPC. In the second step, to properly alleviate the overenhancement artifacts, the proposed method adjusts the degree of contrast improvement in the mapping functions based on a global attribute that represents the contrast of the input image. Furthermore, the proposed method modifies the mapping functions using a local attribute, the spatial correlation between the input and output images, to alleviate the spatial distortion. Finally, using the modified mapping functions, the output image is generated to accurately meet the TPC. In the experiments, the proposed method increased the structural fidelity, enhancement performance measure, and sharpness by up to 13.1%, 65.7%, and 44.7% compared to the benchmark methods. Furthermore, it reduced the execution time by up to 28.1% and 91.6% compared to the benchmark methods and noniteratively met the TPC within an error rate of approximately 0.43%.
- URI
- https://oasis.postech.ac.kr/handle/2014.oak/37295
- DOI
- 10.1109/JDT.2016.2595587
- ISSN
- 1551-319X
- Article Type
- Article
- Citation
- Journal of Display Technology, vol. 12, no. 11, page. 1257 - 1267, 2016-11
- Files in This Item:
- There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.