| Author | Tran Minh Thuan |
| Call Number | AIT Thesis no.ISE-06-35 |
| Subject(s) | Robots
|
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering |
| Publisher | Asian Institute of Technology |
| Abstract | A robot is engineered by a fusion of fields of science and techniques. This thesis describes a novel solution to implement the vision ability for a human-following robot which is able to recognize and locate a nominated human. We use shape-based recognition approach. Initially, the -robot builds a database of information, or knowledge, about the human who will be followed. In this stage, background subtraction and graph-based image segmentation do identify the most meaningful parts of the human. Visual features of these parts and the relationship between them are analyzed to describe that human. After holding the information, the robot is ready to look for the nominated human in captured images in real-time. Recognition filters which are depth filter, color filter, shape filter and relationship filter were created to support the human recognition. Stereo-vision constructs depth maps that are used to remove objects far away from the robot and to locate the position of the detected target in the scene. Experimental results demonstrated the feasibility and robustness of our solution and its capability to deal with arbitrary environments |
| Year | 2006 |
| Type | Thesis |
| School | School of Engineering and Technology (SET) |
| Department | Department of Industrial Systems Engineering (DISE) |
| Academic Program/FoS | Industrial Systems Engineering (ISE) |
| Chairperson(s) | Afzulpurkar, Nitin V.; |
| Examination Committee(s) | Phan Minh Dung;Manukid Parnichkun; |
| Scholarship Donor(s) | AIT Fellowship; |
| Degree | Thesis (M.Eng.) - Asian Institute of Technology, 2006 |