Performance Optimization of RouteSegmentation Algorithm Using Douglas-Peucker Line Simplification Approach

RouteSegmentation had not been tested for intercity travel with far more route points (up to 2,000), where processing cost rises sharply. This study applies Douglas-Peucker line simplification (a 50-meter threshold) before running RouteSegmentation, to cut the point count without sacrificing the route's visual shape.

For intercity routes with thousands of points, RouteSegmentation becomes far slower and more memory-hungry, since the number of polygons that must be built and merged grows proportionally with the point count.

Douglas-Peucker simplifies the route line by discarding points that deviate less than 50 meters from the straight line between significant points, before the remaining points are processed by RouteSegmentation. The result is compared against running RouteSegmentation directly without simplification.

Tested on a 2,000-point route across East Java, split into 20 subsets, comparing point count, processing time, and memory usage with and without the Douglas-Peucker step.

Douglas-Peucker cut the point count by an average of 96.84%, significantly reducing RouteSegmentation's processing time and memory usage without a visible drop in the quality of the resulting perimeter area (Mann-Whitney test, p<0.001).

Demonstrates that Douglas-Peucker line simplification as a pre-processing step makes RouteSegmentation viable for large-scale intercity routes, not just intracity ones.

The simplification threshold (50 meters) is fixed and its sensitivity to other threshold values was not tested; garbage collector behavior again affected memory-measurement accuracy.