DocumentCode
118289
Title
VGEF: Contrast enhancement of dark images using value gap expansion force and sorted histogram equalization
Author
Youngbae Kim ; Jin-Hwan Kim ; Chang-Su Kim
Author_Institution
Sch. of Electr. Eng., Korea Univ., Seoul, South Korea
fYear
2014
fDate
9-12 Dec. 2014
Firstpage
1
Lastpage
4
Abstract
We propose a novel contrast enhancement method for dark images using the value gap expansion force (VGEF) and the sorted histogram equalization. Based on the observation that the inter-pixel relationship is analogous to the electrostatic force, we define the pixel field spread around each pixel and the pixel mass at each pixel position. We compute the VGEF exerted to a pixel by multiplying the pixel field and the pixel mass. Also, we sort the pixels of the same value into 5 clusters according to their VGEF magnitudes to reduce contour artifacts in the enhanced image. Finally, we obtain output pixel values using the transformation function that equalizes the sorted histogram. Experimental results demonstrate that the proposed algorithm effectively enhances the contrast of dark images, and is applicable to the performance improvement of the saliency detection as well.
Keywords
image enhancement; object detection; VGEF; contour artifact reduction; dark image contrast enhancement; electrostatic force; interpixel relationship; pixel field; saliency detection performance improvement; sorted histogram equalization; value gap expansion force; Accuracy; Clustering algorithms; Electrostatics; Force; Histograms; Image segmentation; Visualization;
fLanguage
English
Publisher
ieee
Conference_Titel
Asia-Pacific Signal and Information Processing Association, 2014 Annual Summit and Conference (APSIPA)
Conference_Location
Siem Reap
Type
conf
DOI
10.1109/APSIPA.2014.7041718
Filename
7041718
Link To Document