PPOS SYSTEM: A System of Partitioning Polygonal Objects
Jin Jian-guo · 2009
Partitioning polygonal object into meaningful parts is one of the key steps to recognize this kind of object. Psychologist thinks that when people recognize object, they first decompose object recursively into several parts to recognize, and then combine the parts as a whole to know what the object is. So how to partition object is an important research point in Pattern Recognition and Computer Graphics. In this paper, we introduce our PPOS system, which is a system of partitioning polygonal objects and evaluating the result in a quantitative way. In the paper, we describe the components of our system, and introduce our method to realize them. The system has 3 main parts: partitioning objects, training parameters which are used for partition, and testing the partition result. The test in a quantitative way shows the cuts of our PPOS system fit human's cuts well.