An efficient construction algorithm for binary interval tree nodes

The preprocessing almost take half of the whole time in isosurface extraction of large-scale data, thus it will make sense to shorten the preprocessing time to improve visualization efficiency. We develop a new construction algorithm for binary interval tree nodes based on partitioning span space, and it reduce the sorting computation considerably in isosurface extraction preprocessing. Both complexity analysis and experiments demonstrate the effectiveness of our new technique, and it has 50% around performance promotion in our test data sets. So the new algorithm is practical in large-scale data isosurface extraction, especially for the real-time visualization system.