In this paper, we proposed a method for measuring surface reflectance under multiple lighting by using simple devices. In order to establish this method, we first applied principal component analysis (PCA) to MERL BRDF database for compact representation. For estimating BRDF under multiple lighting, lighting component is obtained from the image captured by omni-directional camera. Through the optimization process using these basis and lighting component, an appropriate set of vertex data necessary for estimating unknown BRDF are selected. By performing the optimization process, our method allows us to estimate unknown BRDF efficiently. Finally, to obtain weighting coefficient for linear combination, the measured data of selected vertex are projected onto the basis of BRDF. We verified that our method can estimate the surface reflectance with high accuracy by comparing the results between ground truth and estimated data.
Yuto Hirasawa, Shoji Yamamoto, Junki Yoshii, Kensuke Fukumoto, Hiroshi Kintou, Norimichi Tsumura, "BRDF Estimation with Simple Measurement and Data-driven Model" in Proc. IS&T 26th Color and Imaging Conf., 2018, pp 371 - 376, https://doi.org/10.2352/ISSN.2169-2629.2018.26.371