International Journal of Applied Information Systems |
Foundation of Computer Science (FCS), NY, USA |
Volume 5 - Number 4 |
Year of Publication: 2013 |
Authors: Ajit Danti, Jyoti Y. Kulkarni, P. S. Hiremath |
10.5120/ijais12-450894 |
Ajit Danti, Jyoti Y. Kulkarni, P. S. Hiremath . A Realtime Road Boundary Detection and Vehicle Detection for Indian Roads. International Journal of Applied Information Systems. 5, 4 ( March 2013), 25-35. DOI=10.5120/ijais12-450894
Traffic conditions in Indian urban and sub urban roads are in many ways not ideal for driving. This is due to faded and unmaintained lane markings. Therefore driving sometimes becomes difficult. Due to inappropriate markings of the roads, it is difficult to track the lane marking using conventional lane marking algorithms. Therefore the issue of Lane tracking with road boundary detection and other vehicle tracking for Indian road conditions is addressed here. The technique is based on modified road boundary detection which first segments the road area based on color segmentation and Hough transform is applied to find out the near vertical lines. Even in the absence of prominent lanes in the road, the segmentation line itself acts as boundary line. Further optical flow based vehicle detection is integrated with the system. When compared with conventional hough transform based lane detection this technique is proven to be more efficient in terms of accuracy. The method is tested with OpenCV under real time environment with Live Video frames. Results show accurate detection of road boundary, lanes and other vehicles under different road textures and varying intensity conditions.