P-33: Color Classification of Images using Categorical Color Database

Abstract

Color classification using a pixel distribution criteria based on the categorical color database was done, and the adequateness of the criteria was examined by color naming experiment. Fifteen color categories were employed. For the category groups of "red", "green", and "blue", the criteria as well as the categorical database look fairly good, but still to be modified. Observers tend to label color names of main object in the image regardless of the pixel percentage. For some category groups, modification of categorical color database seems to be necessary.

53-1:Invited Paper: The Winter Green and Summer Blue

“…Fifteen color images are selected for the test stimuli. They are representatives of red, pink, orange, yellow, green, blue, purple, red-green, red-blue, yellow-green, yellow-blue, blue-green, skin, multi-colors, and monochromatic groups according to the criteria based on our trial and error [8,9]. In Exp.2, one greenish image is added.…”

Threshold Metric Chroma of Images for Chromatic Perception

“…Nine different images representing the following color names, "red", "green", "orange", "yellow", "pink", "blue", "purple", "red and green", and, "yellow and blue", were chosen as original images. Color names of each image were based on our previous study [6]. Metric chroma Cab * was calculated for each pixel of original image and then multiplied by multiplying factor k from 0 to 0.2.…”