Intelligent Area Segmentation of Large Areas for Optimized UAV Flight Path Planing with Minimal Sharp Turns

Yeap, Zhi Teng (2025) Intelligent Area Segmentation of Large Areas for Optimized UAV Flight Path Planing with Minimal Sharp Turns. Final Year Project (Bachelor), Tunku Abdul Rahman University of Management and Technology.

Text Yeap Zhi Teng - FULL TEXT.pdf Restricted to Registered users only Download (5MB)

Abstract

This project presents an intelligent area segmentation algorithm designed to optimize Unmanned Aerial Vehicles (UAVs) flight paths over large agricultural estates. Traditional UAVs flight paths often include sharp turns that will lead to inefficient operations, increased energy consumption, and reduced flight time. Synthetic estate maps were generated using Malaysian state outlines, and three segmentation methods, K-Means Clustering, Agglomerative Hierarchical Clustering (AHC), and Area Segmentation with Axis Projection (ASAP) were applied. The segmented areas will be evaluated using two search algorithms, Smart Greedy Nearest-Neighbour and Greedy Insertion, to determine the most efficient UAV flight paths. Results showed that the ASAP method produced the most elongated segments, which having the largest elongated ratio 9.7083. This helps to smoother UAV paths with fewer sharp turns. Future improvements include applying the real-world datasets and integrating environmental factors for adaptive, real-time UAV path planning.