Intrinsic scene properties from hyperspectral images and LiDAR

In this paper, a novel reflectance model is proposed to recover intrinsic images from remote sensing hyperspectral images (HSIs). Intrinsic image recovery is a well-known challenging and underconstrained problem in computer vision and it becomes even more severely ill posed for HSIs. To reduce the uncertainties and improve the recovery accuracy, two kinds of priors are introduced: 1) shading prior which describes the geometric relation between illuminate and object surface; 2) reflectance prior based on L1-graph coding, which describes the relation between pigment density with reflectance. These priors can effectively eliminate the reflectance inhomogeneity caused by surface normal changes or pigment density variations other than material changes. Then, a non-iterative optimization method is proposed to combine the shading prior and reflectance prior, with which closed-form solutions can be derived and thus avoided falling into local optimums. The experimental results demonstrate that the proposed method can efficiently improve the spectral reflectance homogeneity within a class while preserving the image boundaries; it also produces competitive performance with the state-of-art when utilizing the extracted intrinsic hyperspectral reflectance feature in the task of HSI classification.