Interpolating and compressing volume data with MPU
Seiji Inoue, Kazuyuki Kojima, Susumu Nakata, Satoshi Tanaka · Journal of the Visualization Society of Japan · 2007
Volume visualization is widely used in medical, scientific, and other fields. Since volume data are generally large, data compression techniques become important. In addition, volume data are usually discrete, and must be interpolated to execute volume-visualization algorithms. For high-quality visualization, higher-order smooth interpolation is required. But it is difficult to make such interpolation compatible with the data compression. Recently, the MPU (multi-level partition of unity) method has been developed to reconstruct a smooth implicit surface from a large number of 3D points. Our purpose is to construct a new version of the MPU method that is applicable to interpolate discrete volume data, generating a highly-smooth scalar field described with small-size data. It is demonstrated that the data size is reduced drastically, keeping the smooth feature of the interpolated scalar field.