2007 Volume 46 Issue 5 Pages 357-362
Physical measurements on gold color object can only indicate that the color is yellow. Recognition mechanism for gold color has not been sufficiently clarified yet. In this report, we propose a hypothesis that recognition of gold color requires recognition of shape of the object as a necessary step before the recognition of gold color. Experiments were performed for the purpose of verifying our hypothesis. Recognition of object shapes and recognition of gold color were evaluated on sample images of various sharpness from an extremely low level to a high level;sample images are photographs of gold objects. Two other experiments were also performed using images of various resolutions and using various showing time of images. All the results showed that recognition of gold color was always led by recognition of object shapes. Our hypothesis, which is supported by our experimental results, suggests that gold color should not be categorized in the group of other ordinary colors for instance red, green, blue, cyan, magenta, yellow and so on, and that recognition of gold color consists of color recognition of yellow and recognition of its spatial distribution, which means a shiny surface of the object. Recognition mechanism for silver color can be explained the same way as that for gold color.
