This is a study of utilizing a black-and-white (b/w) image input device to obtain any color image by applying the knowledge of additive color system and color image processing. The main objective was to generate a color image form three sets of intensity data of its primary color components (red, green, and blue – RGB) which were acquired by capturing color pictures using a b/w image input device with a set of RBG filers applied. The process was performed separately for each filter. Since each selected filter had its own limitation and characteristics, some techniques of image processing. Especially histogram modification, were then used to calibrate the data before the color mixing was performed. Furthermore, a software tool for assigning pseudo colors to a gray scale image and a software tool for adjusting brightness and contrast of a color image were developed. This research resulted in a development of Black-and-White to Color (BW2COLOR) software, a software to compose color images with a b/w image input device. The output color images were obtained by capturing color images with a b/w scanner applied with RGB cellophane filters, calibrating data sets with four gray bands on a calibration card, and composing them with the BW2COLOR software. In order to measure the quality of the output color images, a measurement scheme was designed. A software tool to measure color distortions was also developed and used in this research. Five samples of color images resulted from this method were compared to the color images obtained from a color scanner. The results reveal an average normalized mean error (NME) of 20.86% with an average percentage of root mean square error per mean error (erms/ME percentage) of 107.83%. This research was supported by National Science and Technology Development Agency. A paper describing this research was published in National Computer Symposium Proceedings, 1994.
