Spectral Reflectance Reconstruction of Organic Tissue Based on Camera Responses

Reflectance information of a tissue is desired in the field of medical imaging, especially for disease diagnosis. In this paper, a new reflectance restoration algorithm is proposed to recover the reflectance information using a commercial camera. Initially, a color clustering method was applied to obtain the representative colors of tissue samples. These colors were then used to construct a look-up-table (LUT) using a lattice regression model. Interpolation methods can then be applied to the newly built LUT to acquire its matching reflectance information for any RGB input images. Present results showed that the proposed method further improve the accuracy in spectral reconstruction for tissue samples.