Autonomous UAV Target Search Method Based on Lightweight YOLOv8n and Coverage Path Planning.

Area of Science:

• Robotics
• Computer Vision
• Artificial Intelligence

Background:

• Unmanned Aerial Vehicles (UAVs) are crucial for tasks like disaster search and rescue, but face computational and coverage limitations.
• Efficient autonomous target search is a key challenge for UAVs in complex environments.

Purpose of the Study:

• To develop an autonomous UAV search method combining lightweight target detection and coverage path planning.
• To address computational constraints and improve environmental coverage efficiency for UAVs.

Main Methods:

• Lightweight target recognition using pruned and quantized YOLOv8n with HSV data augmentation.
• Dual-layer path planning: spatial coverage via Breadth-First Search (BFS) with greedy optimization and A* for backtracking.
• Construction of a grid-based search environment model.

Main Results:

• The proposed method significantly enhances UAV target search efficiency and reduces path redundancy.
• Recognition accuracy is maintained even in complex environments.
• The approach proves effective for autonomous UAV search on resource-constrained platforms.

Conclusions:

• The integrated approach offers an effective solution for autonomous UAV target search.
• Optimized target detection and path planning improve mission performance.
• This method enables efficient autonomous operations for UAVs with limited onboard resources.