Ranged Filtering of Streaming Numeric Data, or Geolocation Filtering of Streaming GPS Data, using Topic-Based Pub/Sub Messaging

摘要

This paper describes an innovative method to implement location-based filtering of streaming GPS point data without the use of a geospatial database or GIS engine. The filtering algorithm takes a set of polygons as input, and generates an approximating shape of a specified relative accuracy as output, consisting of a union of rectangles. These rectangles can be used to perform point-in-polygon comparisons quickly using text-based pattern matching in a pub/sub messaging system, which excels at handling streaming data at scale. This method has applicability in transportation, logistics, smart cities, security, and surveillance, and can be generalized to one or three dimensions.