Ellipsoidal ray casting algorithm

The Digital Earth (DE) offers a visualization platform via a reference ellipsoid (RE). However, for global volume data on the DE, ray casting algorithm (RCA) faces the challenges of poor performance, earth curvature and incorrect volume rendering due to ignoring earth occlusion. To address the issues, RCA is refactored as ellipsoidal ray casting algorithm (ERCA), which emphasizes the following aspects: ellipsoid- texture space logic mapping, ellipsoidal trilinear interpolation, ray sampling optimization operators, distinguishable transfer function, and proxy ellipsoidal geometry. In particular, a new curved multi-scale ellipsoidal volume grid (EVG) is proposed, which uniformly represents local or global volume data and supports logical mapping between ellipsoidal space and texture space. Ray sampling is optimized with a novel set of culling operators: proxy geometry culling, earth occlusion culling and ellipsoidal bounding box culling, which remove a large number of invalid sampling points to ensure high performance and eliminate correct rendering of ERCA. Experimental results show that ERCA is valid. The ellipsoid-based volume rendering framework can help humans gain insight into 3D scalar data overlaid on the DE.