For Colorization using Template Matching Basic Research on
Kenji Sakoma1, Makoto Sakamoto2
1Kenji Sakoma, Department of Computer Science and System Engineering, University of Miyazaki, Japan.
2Makoto Sakamoto*, Department of Computer Science and System Engineering, University of Miyazaki, Japan.
Manuscript received on August 01, 2020. | Revised Manuscript received on August 05, 2020. | Manuscript published on September 30, 2020. | PP: 517-519 | Volume-9 Issue-3, September 2020. | Retrieval Number: 100.1/ijrte.C4628099320 | DOI: 10.35940/ijrte.C4628.099320
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Colorization, also known as colorization, is a term introduced by Wilson Markle in 1970, and is a method of coloring black-and-white images and videos using a computer. Coloring is important. Imagine coloring a picture as an example. The painting before painting only gives information of existence, such as trees, flowers, and clouds. Some things can be identified by color. This does not give us enough information from the picture. But what about coloring? If you paint the sky red, it will be a sunset, and if you paint the ground green, it will be a meadow. In other words, it is possible to express not only the background but also the background. This makes it possible to read information that cannot be understood only in black and white. With the development of digital devices such as smartphones these days, the chances of seeing black-and-white images are decreasing, but in modern times, black-and-white images are used for X-ray images, MRI images, aerial photographs, fixed-point observations, etc. There are many opportunities to be lost. The development of color photography began in the world in the 1800s, and the development of color photography began in Japan in 1940. In other words, the photographs before that were black and white, and colorization was used to colorize them. Currently, many researchers are studying colorization methods and processes, and the processing time and the burden on users are being reduced. However, software that can perform highly complete colorization is expensive, and some are complicated to operate. Therefore, in this research, as basic research for the development of a fully automatic colorization program for free software, color images (template images) lacking information by template matching using ZNCC and black-and-white images with similar brightness patterns are colored. We made a prototype of a colorization program that restores images.
Keywords: Computer graphics, Colorization, Zero-means Normalized Cross-Correlation, BMP image, HSV color space, Color Propagation.