Modeling Bidirectional Reflectance Factor of Complex Scene Using Iterative MapReduce

Previous Bidirectional Reflectance Factor (BRF) modeling approaches are not sufficient to handle complex scenes, which are radiation-complicated and data-massive. Using iterative MapReduce framework, this paper presents a series of algorithms executable in multiple nodes. Firstly, a virtually scenario is established by 3ds-Max, reconstructable geometry shadow maps method is then employed to count the visibility of patch sets. In addition, endmember variability is accounted for further enhancing simulation fidelity. Finally, an experimental validation is performed. Results by the proposed methods are more coherent with the observation data.