Warm-cool color-based high-speed decolorization: an empirical approach for tone mapping applications

Abstract

Grayscale images are fundamental to many image processing applications like data compression, feature extraction, printing and tone mapping. However, some image information is lost when converting from color to grayscale. In this paper, we propose a light-weight and high-speed image decolorization method based on human perception of color temperatures. Chromatic aberration results from differential refraction of light depending on its wavelength. It causes some rays corresponding to cooler colors (like blue, green) to converge before the warmer colors (like red, orange). This phenomena creates a perception of warm colors "advancing" toward the eye, while the cool colors to be "receding" away. In this proposed color to gray conversion model, we implement a weighted blending function to combine red (perceived warm) and blue (perceived cool) channel. Our main contribution is threefold: First, we implement a high-speed color processing method using exact pixel by pixel processing, and we report a 5.7× speed up when compared to other new algorithms. Second, our optimal color conversion method produces luminance in images that are comparable to other state of the art methods which we quantified using the objective metrics (E-score and C2G-SSIM) and subjective user studies (decolorization and tone mapping). Third, we demonstrate that an effective luminance distribution can be achieved using our algorithm by using global and local tone mapping applications.

Decolorization based on the weighted combination of image entropy and canny edge retention ratio

“…Decolorization aims to preserve as much information (structural features and chrominance contrast) of the original color image as possible. Many decolorization methods 2 – 27 have been proposed to transform color images into grayscale images. They can be classified into two categories: traditional methods (global mapping, 2 – 16 local mapping, 17 – 20 and hybrid methods 21 – 23 ) and deep learning methods 24 …”

“…Many decolorization methods 2 – 27 have been proposed to transform color images into grayscale images. They can be classified into two categories: traditional methods (global mapping, 2 – 16 local mapping, 17 – 20 and hybrid methods 21 – 23 ) and deep learning methods 24 – 27 …”

“…The grayscale value g of any color pixel is a linear combination of InormalR, IG, and InormalB (values of R, G, and B channels), g=αnormalRInormalR+αnormalGInormalG+αnormalBInormalB with fixed weights (αnormalR,αnormalG,αnormalB)=(0.2989,0.5870,0.1140) for all color images to obtain their grayscale results. Some linear global mapping methods 4 – 7 , 15 are similar to rgb2gray(), e.g., Nafchi et al 6 . present a low run-time global mapping method for color-to-gray conversion by estimating the three global linear weighting coefficients.…”

“…There have been no more deep learning-based approaches for color-to-gray conversion in the recent three years. Some traditional methods 7 , 14 – 16 have been proposed for color-to-gray conversion. Li and Zhu 7 present an image decolorization approach based on image clustering and weight optimization.…”

“…It is adaptable to feature preserving, whereas its execution time is not in the leading position compared to other methods. Ambalathankandy et al 15 . proposed a lightweight and high-speed color-to-gray conversion based on human perception of color temperatures.…”