Synthetic Generation of Spatial Polygons On Cloud

Spatial Data generated from varied resources are collected for applications in analysis of land-patterns and population growth among many. While the data collected is vast, majority of it is unavailable in formats or quantity required for research projects. Current data generation tools produce overly rigid geometric objects which do not reflect real-world spatial variations of polygons. We present a web based synthetic generator tool for spatial polygons capable of generating an impressive 0.7 million polygons which would take a range of user inputs including spikiness, irregularity and dataset size to generate requirement quenching data. Generated polygons are thus subject to rigorous benchmarking to ensure that they are close to real-world in addition to being geometrically valid. We further explore the computational problems and propose a spark-based approach of parallel computation, resulting in a remarkable 50% reduction in processing time.