Title page for etd-0809113-164708


[Back to Results | New Search]

URN etd-0809113-164708
Author Yun-shiou Tsai
Author's Email Address No Public.
Statistics This thesis had been viewed 5568 times. Download 627 times.
Department Mechanical and Electro-Mechanical Engineering
Year 2012
Semester 2
Degree Master
Type of Document
Language zh-TW.Big5 Chinese
Title Real-time obstacle detection based on stereo vision
Date of Defense 2013-07-24
Page Count 70
Keyword
  • stereo vision
  • SGBM algorithm
  • obstacle detection
  • fuzzy theory
  • Abstract In recent years, the technology advances very fast. The research of advanced vehicle control and safety system (AVCSS) becomes more critical in the world. AVCSS includes sensor, computer, communication and control technologies applied on the vehicle for assisting driving safety. The core technology consists of a vehicle anti-collision system, automatic driving, and automatic parking. However, these technologies must have the image information in front of vehicle.
    The machine vision can be divided into monocular vision and binocular stereo vision. The environmental information provided by monocular visual images may not be sufficient for the safety requirements upgrade. Besides, the binocular stereo vision can get more ambient information. By stereo vision algorithms, the system can calculate the depth image information to infer depth map information. The image matching algorithm is also important in stereo vision. Therefore, this thesis proposes an obstacle recognition system based on semi-gobal matching algorithm (SGBM) which used fuzzy theory to select a threshold to achieve better recognition results.
    Advisory Committee
  • Yih-tunTseng - chair
  • Pei-Hsuan Tsai - co-chair
  • Jau-Wei Perng - advisor
  • Files
  • etd-0809113-164708.pdf
  • Indicate in-campus at 5 year and off-campus access at 5 year.
    Date of Submission 2013-09-09

    [Back to Results | New Search]


    Browse | Search All Available ETDs

    If you have more questions or technical problems, please contact eThesys