26 July 2007 Polygon cluster pattern recognition based on new visual distance
Author Affiliations +
The pattern recognition of polygon clusters is a most attention-getting problem in spatial data mining. The paper carries through a research on this problem, based on spatial cognition principle and visual recognition Gestalt principle combining with spatial clustering method, and creates two innovations: First, the paper carries through a great improvement to the concept---"visual distance". In the definition of this concept, not only are Euclid's Distance, orientation difference and dimension discrepancy comprehensively thought out, but also is "similarity degree of object shape" crucially considered. In the calculation of "visual distance", the distance calculation model is built using Delaunay Triangulation geometrical structure. Second, the research adopts spatial clustering analysis based on MST Tree. In the design of pruning algorithm, the study initiates data automatism delamination mechanism and introduces Simulated Annealing Optimization Algorithm. This study provides a new research thread for GIS development, namely, GIS is an intersection principle, whose research method should be open and diverse. Any mature technology of other relative principles can be introduced into the study of GIS, but, they need to be improved on technical measures according to the principles of GIS as "spatial cognition science". Only to do this, can GIS develop forward on a higher and stronger plane.
© (2007) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Yun Shuai, Haiyan Shuai, Lin Ni, "Polygon cluster pattern recognition based on new visual distance", Proc. SPIE 6753, Geoinformatics 2007: Geospatial Information Science, 675316 (26 July 2007); doi: 10.1117/12.761778; https://doi.org/10.1117/12.761778


A topological-based spatial data clustering
Proceedings of SPIE (April 20 2016)
Visualizing research of land use land cover change
Proceedings of SPIE (November 05 2008)
A perspective of spatial data mining
Proceedings of SPIE (December 02 2005)

Back to Top