The physical-based geometric-optical Li-Strahler model can be inverted to retrieve forest canopy structural variables. One of the main input variables of the inverted model is the fractional component of sunlit background (K g). K g is calculated by using pure reflectance spectra (endmembers) of the viewed surface components. In this paper, the feasibility of up-scaling from high (Quickbird) to medium (Hyperion) spatial resolution data for extracting the required endmembers is demonstrated. Furthermore, the sensitivity of the endmembers used as input for inverting Li-Strahler model is evaluated. After validating the inverted model results, namely spatially explicit forest mean crown closure and crown diameter using field measurements, it can be concluded that the regional scaling-based endmembers derived from the linear unmixing model are the best ones to be used in combination with the inverted Li-Strahler model for quantitatively monitoring disturbance in forest canopy structure.