Outdoor Terrain Classification Algorithms Using SAM and YOLO

Institut

Lehrstuhl für Agrarmechatronik (WZW)

Typ

Masterarbeit /

Inhalt

experimentell / theoretisch /  

Beschreibung

In the field of outdoor robotics, autonomous navigation, and environmental monitoring, accurate terrain classification is essential for informed decision-making. Identifying and understanding various terrain types and obstacles, such as grass, dirt, water bodies, and obstacles, is crucial for ensuring the safe and efficient operation of autonomous systems. This project aims to develop terrain classification algorithms using the innovative Segment Anything Model (SAM) in conjunction with You Only Look Once (YOLO) object detection to enhance the perception capabilities of autonomous systems in outdoor environments.

Voraussetzungen

The theoretic part of the thesis can be done completely in home office. For the field experiments, the student should be able to come to the chair to Dürnast (close to Weihenstephan). A general understanding of machine learning specially Computer Vision is beneficial. Structured and independent work is anticipated. The thesis is intended as Master’s thesis, but can be adapted to fit different thesis needs on Master’s level.

Möglicher Beginn

sofort

Kontakt

Ausschreibung