City University of Hong Kong
DSpace
 

CityU Institutional Repository >
4_Student Final Year Projects >
Computer Science - Undergraduate Final Year Projects >

Please use this identifier to cite or link to this item: http://hdl.handle.net/2031/471

Title: Project AIBO - machine learning and object recognition
Authors: Shum, Tsz Kin
Department: Department of Computer Science
Issue Date: 2004
Supervisor: Prof. Ip Horace H S. First Reader: Wong Hau San. Second Reader: Chun H W Andy
Abstract: AIBO in Japanese means “companion”. It is also an acronym for Artificial Intelligence Robot. AIBO expresses real emotions and instincts. Explore the world of AIBO and discover a friendship that matures with time. AIBO is a robotic dog developed by SONY, it contained various sensor including touch sensor, temperature sensor, camera, distance and sound sensor, mic and LEDs. With support of RISC CPU and real-time OS, AIBO can be programmed to be have intelligent and do different kind of motion as well as tasks. AIBO provided us a chance to explore topics on AI, multi-agent system, computer vision, sound recognition, strategy acquisition, real-time reasoning and more. Recently, robocup is being held and use AIBO to play football match. It is a task for a team of multiple fast-moving robots under a dynamic environment. Its ultimate goal is to develop a fully autonomous humanoid robots that can win against the human world champion team in soccer. It is not difficult to imagine that robots will soon enter our life and help us to finish some daily tasks. We will work close together will robot in the future, just like the scenes in the movie A.I. As robot will become more and more important to human life, it is worth to do research on various topics about robotics. In this report, we define computer vision as the main topic. The technique to achieve computer vision and the methods on object recognition will be discussed. We may put some emphasis on the robotic vision since we will use AIBO to recognize some objects in a specified environment. Two approaches are discussed and tested to proof their feasibility on shape recognition. With support of experiment results, the second approach, using shape area and perimeter to area ratio is used and have implemented on AIBO. This approach can successfully identify some colored objects but not all. The reasons behind will be discussed. For more, difficulties on robotic vision recognition will also be investigated.
Appears in Collections:Computer Science - Undergraduate Final Year Projects

Files in This Item:

File SizeFormat
fulltext.html164 BHTMLView/Open

Items in CityU IR are protected by copyright, with all rights reserved, unless otherwise indicated.

 

Valid XHTML 1.0!
DSpace Software © 2013 CityU Library - Send feedback to Library Systems
Privacy Policy · Copyright · Disclaimer