Open Journal Systems

DWSI: AN APPROACH TO SOLVING THE POLYGON INTERSECTION-SPREADING PROBLEM WITH A PARALLEL UNION ALGORITHM AT THE FEATURE LAYER LEVEL

JUNFU FAN, CHENGHU ZHOU, TING MA, MIN JI, YUKE ZHOU, TAO XU

Abstract


A dual-way seeds indexing (DWSI) method based on R-tree and the Open
Geospatial Consortium (OGC) simple feature model was proposed to solve the polygon intersection-spreading problem. The parallel polygon union algorithm based on the improved DWSI and the OpenMP parallel programming model was developed to validate the usability of the data partition method. The experimental results reveal that the improved DWSI method can implement a robust parallel task partition by overcoming the polygon intersection-spreading problem. The parallel union algorithm applied DWSI not only scaled up the data processing but alsospeeded up the computation compared with the serial proposal, and it showed a
higher computational efficiency with higher speedup benchmarks in the treatment of larger-scale dataset. Therefore, the improved DWSI can be a potential approach to parallelizing the vector data overlay algorithms based on the OGC simple data model at the feature layer level.

Keywords


Dual-way Seeds Indexing Method; Polygon Intersect-spreading; Parallel Union; Task Partition