Abstract
(Introduction) The endoscope is used in not only the diagnosis but also the treatment in every medical field and the use frequency is increasing remarkably. The data volume increases, the burdens to endoscopists increase. (Purpose) We have proposed that the diagnosis support system to reduce this burden. The endoscopists diagnose from color information. However even the same tissue is seen as different color of the mucosa when viewed from far distant and near view. But it is a change of brightness, color information should be the same. However, it is a change of the brightness, and the color information should be the same. Therefore we paid our attention to HSV coordinates which distributed color (hue: wavelength of the light) and brightness (value) definitely. Then we confirmed the color reliability of HSV coordinates. (Method) The endoscopic image is displayed as RGB image. It is easy for computer to convert into HSV coordinates. The spectral image was calculated by a virtual spectrum from the same image. (Results) The color information of HSV indication was the almost same as at any distance of endoscope position. So we will apply this HSV indication to image processing.
