A proposed methodology for colorizing historical grayscale images: using image processing and artificial intelligence applications

Humanity has a vast wealth of historical grayscale photographs that tell the history of this world. Colorization of historical images began in the 1990s by using professional image processing applications and artificial intelligence. Ten years ago, using artificial intelligence applications based on deep learning expanded to execute restoration and colorization of historical images. However, there are problems related to the quality of production, such as the accuracy of the color tones of historical image elements due to the lack of historical information about the image content in artificial intelligence applications, in addition to the lack of an approved system that codifies the uses of these programs to obtain the required quality. Therefore, this research aims to reach standards and design a production system for coloring old grayscale images that increase the quality of coloring and technical specifications of images for use in printed and visual media. The researcher adopted the experimental descriptive approach to propose a methodology that combines the use of artificial intelligence applications and professional image processing applications to carry out restoration and coloring operations more accurately with the intervention of the human element and to determine the sequential and overlapping production processes and the final specifications of the image files. The researcher recommends cooperation between developers of artificial applications and specialists in image processing to develop these systems, and so between developers and historians to provide artificial intelligence applications with historical information from reliable sources to improve the results of color processing of images.