An improved centroid-based isovolume extraction algorithm for 3D Data Sets rendering

In this paper, an approach named isovolume extraction for 3D Data Sets rendering is presented. This approach extracts and visualizes a model that meets the desired range property values. It extends previous visualization techniques and is very suitable for closed objects extraction. After deeply studying the centroid-based isovolume extraction algorithm proposed by David J. Russomanno et.al, authors present an improved centroid-based isovolume extraction algorithm. It works in five main steps: 1) creating 3D regular volume data; 2) judging a voxel in the isovolume or not, if in, dividing it into six tetrahedrons, else removing it; 3) calculating centroid of each tetrahedron with the improved method; 4) extracting the effective tetrahedrons and subdividing the non-effective tetrahedrons, then repeating step 3) recursively to improve the model precision; 5) collecting unique triangles from the tetrahedron set formed by step 3) and 4), and creating a isovolume model. Finally in MapGIS-TDE platform, with the water content data of LinGang and the underwater pollution concentration data as test samples, the improved algorithm has been verified. The data computation time and model information are measured. The experimental results show that the improved algorithm represents the 3D data more realistic and accurate after the modification.