Quality Enhancement of Color-Embedded- Grayscale Image Using Convolutional Neural Network
Heri Prasetyo, Albi Revlin Bagaskara, Umi Salamah · 2021
The color-embedded-grayscale image aims to transform a color image into its grayscale version. It involves the Discrete Wavelet Transformation (DWT) and chrominance-luminance color space. This technique hides the color information contained in the chrominance component into the DWT transformed of luminance component. The color-embedded-grayscale image can be processed to yield a recovered color image. However, the quality of this recovered image is less acceptable compared to the original version. Thus, this paper improves this recovered image using deep learning approach, i.e. Convolutional Neural Network (CNN). As documented in the experiments, the proposed method is superior in comparisons to the other existing schemes.