Distinction of Glossy Colored Objects Using Gray Level
Kazuo Yamaba and Yoichi Miyake*
Mechanical Engineering Labolatory, 1-2 Namiki, Tsukuba, Ibaraki 305, Japan
*Faculty of Engineering, Chiba University, 1-33-Yayoi, Inage-ku, Chiba 260, Japan
A new measuring apparatus for color images has been developed in order to distinguish glossy colored objects under fluorescent lamp illumination. The color image apparatus is mainly composed of a zoom lens, a mirror box, MOS cameras, a microcomputer and a color image processor. The zoom lens can automatically blur an image by the microcomputer. It is a very effective method for obtaining a blurred image in detecting glossy colored objects. Gloss can be omitted by blurring an original image of objects. In the blurred image, it is known that tristimulus values are not affected by image restoration if the modified Wiener filter method is employed. The blurred image is restored by the filter and is processed by a new analogical method utilizing a fuzzy set theory based on a visual psychophysical method. As a result of this experiment, it can be concluded that the system demonstrates the possibility of highly accurate distinction of glossy colored objects.