Fast image enhancement based on color space fusion

Abstract

The current Retinex algorithm processes the RGB channels separately for color image enhancement. However, it changes the ratios of RGB components and also causes some serious problems, such as color distortion, color noise, and the halo artifacts. To solve these issues, we propose a novel algorithm based on color space fusion. The single scale Retinex with fast mean filtering is applied to the luminance component in hue‐saturation‐value (HSV) color space. An enhancement adjustment factor is introduced to avoid color distortion and noise amplification. Then, the surrounding function is replaced by a small scale Gaussian filter in RGB color space to eliminate the halo artifact. A parameter is involved to keep the color natural when the reflection is estimated. Finally, the enhanced color image is constructed from the weighted averaging results of these two steps. The subjective and objective evaluations of many different backlight images captured by different cameras are implemented to verify the validity of the proposed algorithm in our experiments. The experimental results show that the proposed algorithm can not only significantly suppress the halo effect and noise amplification, but can also remove color distortion. Our proposed algorithm is superior to the multi‐scale Retinex with color restoration approach and other state‐of‐the‐art methods. © 2014 Wiley Periodicals, Inc. Col Res Appl, 41, 22–31, 2016

Change Detection Using High Resolution Remote Sensing Images Based on Active Learning and Markov Random Fields

“…RGB color values are clustered together based on their discriminative power in a classification problem [52], so that each cluster has the explicit objective to minimize the decline of mutual information of the final representation. Besides, this kind of color description can automatically maintain photometric invariance to some extent.…”

Low-Light Image Enhancement Based on Guided Image Filtering in Gradient Domain

“…Li et al [18] proposed the WGIF, where a spatially varying gradient weight constraint 2 ( ) = /Γ was added in (7).…”

“…The commonly used methods include dark channel prior model [1,2], neural network model [3,4], histogram equalization (HE) [5,6], image fusion [7,8], wavelet domain algorithm [9,10], and illuminationreflection model [11][12][13][14]. It is noted that the adaptability of dark channel prior model is poor in disposing the images with rich details and high brightness [2].…”

Image Enhancement for Surveillance Video of Coal Mining Face Based on Single-Scale Retinex Algorithm Combined with Bilateral Filtering

“…Xie et al [14] used guided filter to improve SSR and got good results in enhancing the image quality and finger vein recognition accuracy. Xiao et al [16] employed a fast mean filtering to improve the performance of SSR. Fu et al [17] tried simultaneous estimation of illumination and reflectance in the linear domain to improve Retinex, and their method was proved to be better than other common methods.…”