Ir. H.W. Ho
PhD candidate, working at Department of Control & Simulation, Faculty of Aerospace Engineering at Delft University of Technology. I am currently focusing on the use of computer vision to make the Unmanned Aerial Vehicle more intelligent and autonomous.
Flight Control, Computer Vision, Unmanned Aerial Vehicle, Artificial Intelligent
- Best Graduate Student Paper Award, CEAS EuroGNC in 2013
- USM Gold Medal Award for the Best Final Year Student in Aerospace Engineering in 2009
Intelligent Vision-based Landing System of UAV
The capability of Unmanned Aerial Vehicle (UAVs) to perform landing task is essential for autonomous operation, especially when they are outside line-of-sight or in emergency situation. For autonomous landing, these vehicles need to know by themselves which areas are safe to be landed upon using small and light-weight sensors that can be carried on-board. This brought about the main goal of my research, i.e. to perform autonomous landing of a UAV with a monocular camera using generic computer vision methods. A vision algorithm using optical flow is introduced to compute some important and useful information for landing, such as ventral flow, time-to-contact, surface flatness, and surface slope. These parameters can be used to detect obstacles, search for a safe landing spot, and perform smooth landing. The algorithm is computationally efficient and suitable for use during relatively fast maneuvers.