유사색 모집단을 이용한 개선된 분광 반사율 추정

Advanced surface spectral-reflectance estimation using a population with similar colors

  • 이철희 (경운대학교 컴퓨터공학과) ;
  • 김태호 (대구대학교 정보통신공학과) ;
  • 류명춘 (경운대학교 컴퓨터공학과) ;
  • 오주환 (경운대학교 컴퓨터공학과)
  • 발행 : 2001.05.01

초록

The studies to estimate the surface spectral reflectance of an object have received widespread attention using the multi-spectral camera system. However, the multi-spectral camera system requires the additional color filter according to increment of the channel and system complexity is increased by multiple capture. Thus, this paper proposes an algorithm to reduce the estimation error of surface spectral reflectance with the conventional 3-band RGB camera. In the proposed method, adaptive principal components for each pixel are calculated by renewing the population of surface reflectances and the adaptive principal components can reduce estimation error of surface spectral reflectance of current pixel. To evacuate performance of the proposed estimation method, 3-band principal component analysis, 5-band wiener estimation method, and the proposed method are compared in the estimation experiment with the Macbeth ColorChecker. As a result, the proposed method showed a lower mean square ems between the estimated and the measured spectra compared to the conventional 3-band principal component analysis method and represented a similar or advanced estimation performance compared to the 5-band wiener method.

키워드