Modified Lawn-Mower Search Pattern for Areas Comprised of Weighted Regions

This work focuses on surveillance problems on areas comprised of weighted regions using unmanned aerial vehicles. The objective is to develop a path planning algorithm in order to maximize area coverage and minimize coverage time. Each region of the search area is weighted with respect to other regions of the search area. The weight of a region indicates the uncertainty within the region. Regions with a higher weighting are more important or have a higher uncertainty and hence the vehicle should concentrate more on those regions than on regions with lower weightings. The weights on regions are represented by differences in uncertainty. An uncertainty is assigned to each region. The definition of coverage is introduced as the summation of uncertainty. The goal of the planner is to maximize coverage and to minimize time while satisfying the constraints on the vehicle, such as velocity and turning radius constraints. The planner generates a modified lawn-mower search pattern to achieve the objectives resulting in more than 96% coverage, on average, and reduced search distances by as much as 38% compared to the typical lawn-mower pattern.