Title page for etd-0808115-222352


[Back to Results | New Search]

URN etd-0808115-222352
Author Ching-Hui Wang
Author's Email Address No Public.
Statistics This thesis had been viewed 5341 times. Download 0 times.
Department Computer Science and Engineering
Year 2014
Semester 2
Degree Master
Type of Document
Language zh-TW.Big5 Chinese
Title Lane, Vehicle, and Human Detections in Intelligent Driver Assistance Systems
Date of Defense 2015-07-29
Page Count 88
Keyword
  • HOG
  • Pedestrian Detection
  • Human Detection
  • Driver Assistance Systems
  • Lane Detection
  • Vehicle Detection
  • Abstract In this thesis, we present an integrated driver assistance system with lane detection, vehicle detection, and human detection based on camera image sequences. After image preprocessing, we detect driving lanes, vehicles, and human in the region of interest. First, we propose a simple method to find the traffic lanes. In particular, we identify the traffic lanes of either straight line or curve. Furthermore, before the vehicle detection step, we determine the situation of daytime or nighttime, and then use different methods to detect the front vehicle in the region of traffic lanes. In daytime, we use the shadow detection technique and edge detection to find the regions of cars. In nighttime, pairs of tail lights of vehicles are used to detect vehicles. Finally, we use a popular algorithm, Histograms of Oriented Gradients (HOG), as a feature descriptor combined with Support Vector Machine (SVM) classifier to detect pedestrians.
    Advisory Committee
  • Pei-Yung Hsiao - chair
  • Yun-Nan Chang - co-chair
  • Ming-Chih Chen - co-chair
  • Shen-Fu Hsiao - advisor
  • Files
  • etd-0808115-222352.pdf
  • Indicate in-campus at 5 year and off-campus access at 5 year.
    Date of Submission 2015-09-09

    [Back to Results | New Search]


    Browse | Search All Available ETDs

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