A free-viewpoint application has been developed that yields an immersive user experience. One of the simple free-viewpoint approaches called “billboard methods” is suitable for displaying a synthesized 3D view in a mobile device, but it suffers from the limitation that a billboard should be positioned in only one position in the world. This fact gives users an unacceptable impression in the case where an object being shot is situated at multiple points. To solve this problem, we propose optimal deformation of the billboard. The deformation is designed as a mapping of grid points in the input billboard silhouette to produce an optimal silhouette from an accurate voxel model of the object. We formulate and solve this procedure as a nonlinear optimization problem based on a grid-point constraint and some a priori information. Our results show that the proposed method generates a synthesized virtual image having a natural appearance and better objective score in terms of the silhouette and structural similarity.

We present an algorithm for efficient rendering of animated hair under a dynamic, low-frequency lighting environment. We use spherical harmonics (SH) to represent the environmental light. The transmittances between a point on a hair strand and the light sources are also represented by SH functions. Then, a convolution of SH functions and the scattering function of a hair strand is precomputed. This allows us to efficiently compute the intensity at a point on the hair. However, the computation of the transmittance is very time-consuming. We address this problem by using a voxel-based approach: the transmittance is computed by using a voxelized hair model. We further accelerate the computation by sampling the voxels. By using our method, we can render a hair model consisting of tens of thousands of hair strands at interactive frame rates.

This paper proposes a GPU-based method that can visualize voxelized surface data with fine and complicated features, has high rendering quality at interactive frame rates, and provides low memory consumption. The surface data is compressed using run-length encoding (RLE) for each level of detail (LOD). Then, the loop for the rendering process is performed on the GPU for the position of the viewpoint at each time instant. The scene is raycasted in planes, where each plane is perpendicular to the horizontal plane in the world coordinate system and passes through the viewpoint. For each plane, one ray is cast to rasterize all RLE elements intersecting this plane, starting from the viewpoint and ranging up to the maximum view distance. This rasterization process projects each RLE element passing the occlusion test onto the screen at a LOD that decreases with the distance of the RLE element from the viewpoint. Finally, the smoothing of voxels in screen space and full screen anti-aliasing is performed. To provide lighting calculations without storing the normal vector inside the RLE data structure, our algorithm recovers the normal vectors from the rendered scene's depth buffer. After the viewpoint changes, the same process is re-executed for the new viewpoint. Experiments using different scenes have shown that the proposed algorithm is faster than the equivalent CPU implementation and other related methods. Our experiments further prove that this method is memory efficient and achieves high quality results.

In this paper, we propose a progressive encoding algorithm for binary voxel models, which represent 3D object shapes. For progressive transmission, multi-resolution models are generated by decimating an input voxel model. Then, each resolution model is encoded by employing the pattern code representation(PCR). In PCR, the voxel model is represented with a series of pattern codes. The pattern of a voxel informs of the local shape of the model around that voxel. PCR can achieve a coding gain, since the pattern codes are highly correlated. In the multi-resolution framework, the coding gain can be further improved by exploiting the decimation constraints from the lower resolution models. Furthermore, the shell classification scheme is proposed to reduce the number of pattern codes to represent the whole voxel model. Simulation results show that the proposed algorithm provides about 1.1-1.3 times higher coding gain than the conventional PCR algorithm.

A high speed 3D shape reconstruction method with multiple video cameras and multiple computers on LAN is presented. The video cameras are set to surround the real 3D space where people exist. Reconstructed 3D space is displayed in voxel format and users can see the space from any viewpoint with a VR viewer. We implemented a prototype system that can work out the 3D reconstruction with the speed of 10.55 fps in 313 ms delay.

A support system for hepatectomy that allows segmentation of the liver interactively and directly on 3D images was developed. Intuitive 3D images of the liver and its vessels and tumors were drawn with an improved volume-rendering method. Regions supplied with blood by each branch were interactively identified. 3D segments were defined directly on the images using a mouse and excisions were estimated from these interactive inputs.