ROAD BOUNDARY DETECTION USING SEGMENTATION ON STEREO IMAGES
In this paper, a road detection method based on an image segmentation and stereo vision ispresented. Road detection process is a key issue for an autonomous driving system in urbanenvironment. Image-based road detection algorithm is applied on sources of visual informationrecorded by stereo cameras when our car is running on road. Our method combines a posterioriprobability and visual information for image segmentation. The depth map in stereo camera iscalculated on real time by a circuit board and it is utilized to rectify the boundary on left andright side of road. The method is composed of threesteps. Firstly, a road identifier is trainedwith supervised learning algorithm. Secondly, road regions are detected by combining aposteriori probability and visual information usingimage segmentation algorithm. In the laststep, the segmentation result is combined with the depth-map image to correct the boundary. Experimental results are presented for video sequences of road in urban areas.
Keywords.SWA algorithm, Bayes's rule, depth map, road detection.
Authors who publish with Vietnam Journal of Science and Technology agree with the following terms:
- The manuscript is not under consideration for publication elsewhere. When a manuscript is accepted for publication, the author agrees to automatic transfer of the copyright to the editorial office.
- The manuscript should not be published elsewhere in any language without the consent of the copyright holders. Authors have the right to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal’s published version of their work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are encouraged to post their work online (e.g., in institutional repositories or on their websites) prior to or during the submission process, as it can lead to productive exchanges or/and greater number of citation to the to-be-published work (See The Effect of Open Access).