Extracting Features Using Discrete Wavelet Transform (DWT) and Gray Level Co-occurrence Matrix (GLCM) to Produce Colorization of Images

• Published in: 2023 3rd International Conference on Advancement in Electronics & Communication Engineering (AECE) pp. 711 - 714
• Main Authors: Rengarajan, A., Gupta, Mohan Vishal, Kanna, R. Kishore
• Format: Conference Proceeding
• Language: English
• Published: IEEE 23.11.2023
• Subjects: Discrete wavelet transforms; DWT; Feature extraction; GLCM; Image color analysis; MATLAB; PSNR; Resource management; Textual features; Transforms; Videos
• DOI: 10.1109/AECE59614.2023.10428481

The act of imbuing monochromatic images or videos with hues is commonly referred to as gray scale image colorization. The incorporation of colors not only serves to augment the aesthetic appeal of the image, but also serves to amplify its inherent characteristics. The process of colorization has various applications, including but not limited to medical imaging and scientific illustrations. The process entails the allocation of three-dimensional pixel values to a monochromatic image that solely possesses a single dimension, namely luminance or intensity. Given that distinct colors can possess identical luminance values but differ in terms of hue or saturation, human intervention is typically necessary. However, with the increasing demands of time, it is imperative to reduce the amount of human labour involved. Hence, there is a requirement to devise effective methodologies that facilitate the proficient mapping of colors. The paper outlines a proposed methodology for the implementation of an automated colorization system. The resulting output would exhibit superior performance in terms of evaluation metrics namely PSNR (peak signal to noise ratio) to show colorization quality.