3D Perceptual Organization of Laser Altimetry Data

Perceptual organization is to group sensory primitives arising from a common underlying cause by imposing structural organization on sensory data. It has been emphasized as a robust intermediate-level grouping process toward object recognition and reconstruction since it imparts robustness and computational efficiency to the perceptual process. Sarkar and Boyer (1993) proposed a classificatory structure for perceptual organization and clarified what should be done under each class. Despite intensive research on 2D data, 3D perceptual organization is still in its infancy, however. Increasing research efforts are needed to understand 3D data from various range sensors such as laser altimetry systems. Therefore, the purpose of this research is to develop a robust approach for constructing 3D perceptual organization from irregularly distributed 3D points acquired from laser altimetry systems. The scope of perceptual organization presented in this paper is limited to signal, primitive and structural levels. At the signal level, we organize raw 3D points into spatially coherent surface patches. Then, at the primitive level, we merge the patches into co-parametric surfaces and detect breaklines and occlusions. Finally, at the structural level, we derive useful surface combinations such as polyhedral structures. The approach has been successfully applied to real laser altimetry data. The organized output is on a much more abstract level than the raw data and makes information explicit. Thus, it serves as a valuable input to higher order perceptual processes, including the generation and validation of hypotheses in object recognition tasks.