Spectral Reflectance Reconstruction of Organ Tissue Based on Metameric Black and Lattice Regression

Abstract

In this study, a new approach is proposed for the restoration of reflectance information on organ samples using a commercial camera. This novel approach is comprised of three stages. In the first stage, a color clustering method is utilized to extract the representative colors of the organ samples as well as their corresponding spectral reflectance. In the second stage, the spectral reflectance is decomposed into two separate parts, i.e., the fundamental stimulus spectrum and the metameric black following the matrix-R theory, and the latter is further utilized to form a look-up table (LUT) via a lattice regression model. Finally, the reflectance information can be easily retrieved by referring to the newly built LUT. The performance of the proposed method was investigated, along with that of six other commonly adopted methods, through a physical experiment using real, measured organ samples. The results demonstrate that the proposed method outperformed all the other methods in terms of both colorimetric and spectral metrics, indicating that it is a promising strategy for organ sample reflectance restoration.