Monocular online learning for road region labeling and object detection from a moving platform

  • Authors:

  • Chung-Ching Lin;Marilyn Wolf

  • Affiliations:

  • School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA;School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA

  • Venue:
  • ISVC'11 Proceedings of the 7th international conference on Advances in visual computing - Volume Part II
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

An online learning method is proposed for detecting the road region and objects on the road by analyzing the videos captured by a monocular camera on a moving platform. Most existing methods for moving-camera detection impose serious constraints or require offline learning. In our approach, the feature points of the road region are learned based on the detected and matched feature points between adjacent frames without using camera intrinsic parameters or camera motion parameters. The road region is labeled by using the classified feature points. Finally, the feature points on the labeled road region are used to detect the objects on the road. Experimental results show that the method demonstrates significant object detecting performance without further restrictions, and performs effectively in complex detecting environment.