Navigation System POI Clustering and Route Optimization

The patent segments the POI search and routing process into distinct phases: first identifying multiple POIs independently, then clustering them into groups based on spatial proximity, and finally generating optimized routes between clusters. This segmentation allows the system to handle multiple POIs efficiently by breaking down the complex task into manageable steps.

The patent introduces a spatial dimension by clustering POIs based on their geographic coordinates and proximity to the destination. Instead of treating POIs as a flat list, the system organizes them into hierarchical clusters (first-level clusters near destination, second-level clusters along route), adding spatial structure to the routing problem.

The patent merges multiple POIs that are spatially close to each other into single cluster representatives. Instead of processing each POI individually for routing purposes, the system combines nearby POIs into clusters, reducing the number of routing calculations needed while still providing comprehensive coverage of points of interest.

The system performs preliminary clustering of POIs before generating the final routing. By pre-organizing POIs into clusters based on their proximity to the destination and to each other, the system reduces the computational burden of the subsequent routing optimization, as it only needs to calculate routes between cluster centers rather than between every individual POI.