Autonomous UAV perching on ledge

• Main Author: Ye, Han
• Other Authors: Xie Lihua
• Format: Thesis-Master by Coursework
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Electrical and electronic engineering::Control and instrumentation::Control engineering; Engineering::Electrical and electronic engineering::Control and instrumentation::Robotics
• Online Access: https://hdl.handle.net/10356/158700
• Institution: Nanyang Technological University

Unmanned Aerial Vehicles (UAVs) are playing an ever-increasing role in many fields such as civilian, industrial and military. The development of autonomous technology has made UAVs increasingly intelligent and play an important role in many places where they can replace human power. Despite the many ad- vantages of UAVs, they are often limited by fuel consumption and cannot fly for long periods of time. Based on this situation, we adopt the idea that the UAVs can temporarily land on a roof ledge. By temporarily landing and dock- ing, the UAV can reduce its fuel consumption and increase its usage time and efficiency. This dissertation mainly focuses on the detection part of the perching process, which is the most crucial part of the system. A series of the study was conducted to evaluate the performance of various detection processes such as the multi-plane analysis-based method, the template matching algorithm based on point cloud data, and the deep learning method based on RGB images. The vision-based target detection algorithms can detect potential ledges and thus pro- vide a judgment basis for UAV landing. Practical experiments prove the quality of the approaches.