A Cost-Effective Safe Landing Detection Approach in Unmanned Aerial Vehicle Search and Rescue

Overview

Identifying safe landing zones accurately for Unmanned Aerial Vehicle (UAV) is very challenging, especially during emergency Search and Rescue (SAR) missions. The commonly used LiDAR-based approach is bulky, expensive, and demands high computing power, which makes it not ideal for UAVs. This study presents a computer vision based approach that combines object detection, semantic segmentation, and depth estimation to determine terrain features from 2D RGB images. By integrating these parts, it can locate flat, obstacle free areas suitable for emergency landings in remote environments or areas lacking reliable GPS signals. GPS unreliability limits autonomous navigation and precise landings. This hybrid approach uses low cost software and camera images to achieve a balance of efficiency and cost effectiveness, which makes it more suitable for UAVs, especially smaller drones, where weight, computing power, and budget are limited. The system provides a scalable framework for real time terrain assessment, and can be used in broader emergency response missions. Ultimately, this research creates a foundation for cost effective integration of intelligent landing detection in UAVs used for critical aerial operations.

Competition history

• CSEF 2026 Computational Science (Senior Division) · Entry S-07-09

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• All Computational Sciences projects
• CSEF Computational Sciences projects
• Every CSEF 2026 project