Scene adaptive road segmentation algorithm based on Deep Convolutional Neural Network

被引：7
|
作者：
Wang H. [1 ]
Cai Y. [2 ]
Jia Y. [3 ]
Chen L. [2 ]
Jiang H. [1 ]
机构：
[1] School of Automotive and Traffic Engineering, Jiangsu University, Zhenjiang
[2] Automotive Engineering Research Institute, Jiangsu University, Zhenjiang
[3] Department of Automotive Engineering, Clemson University, 29634, SC
来源：
Cai, Yingfeng (caicaixiao0304@126.com) | 2017年 / Science Press卷 / 39期
基金：
中国国家自然科学基金; 中国博士后科学基金;
关键词：
Auto-encoder; Composite deep structure; Deep Convolutional Neural Network (DCNN); Road segmentation; Scene adaptive;
D O I：
10.11999/JEIT160329
中图分类号：
学科分类号：
摘要：
The existed machine learning based road segmentation algorithms maintain obvious shortage that the detection effect decreases dramatically when the distribution of training samples and the scene target samples does not match. Focusing on this issue, a scene adaptive road segmentation algorithm based on Deep Convolutional Neural Network (DCNN) and auto encoder is proposed. Firstly, classic Slow Feature Analysis (SFA) and Gentle Boost based method is used to generate online samples whose label contain confidence value. After that, using the automatic feature extraction ability of DCNN and performing source-target scene feature similarity calculation with deep auto-encoder, a composite deep structure based scene adaptive classifier and its training method are designed. The experiment on KITTI dataset demonstrates that the proposed method outperforms the existed machine learning based road segmentation algorithms which upgrades the detection rate on average of around 4.5%. © 2017, Science Press. All right reserved.
引用
收藏
页码：263 / 269
页数：6
相关论文
共 17 条
  • [1] Yu T., Wang R., Gu B., Et al., Survey on the vision-based recognition methods of intelligent vehicle road boundaries and lane markings, Journal of Highway and Transportation Research and Development, 38, 8, pp. 139-142, (2006)
  • [2] Zhou H., Kong H., Wei L., Et al., Efficient road detection and tracking for unmanned aerial vehicle, IEEE Transactions on Intelligent Transportation Systems, 16, 1, pp. 297-309, (2015)
  • [3] Shin B.S., Xu Z., Klette R., Visual lane analysis and higher-order tasks: a concise review, Machine Vision and Applications, 25, 6, pp. 1519-1547, (2014)
  • [4] Hillel A.B., Lerner R., Levi D., Et al., Recent progress in road and lane detection: a survey, Machine Vision and Applications, 25, 3, pp. 727-745, (2014)
  • [5] Paz L.M., Pinies P., Newman P., A variational approach to online road and path segmentation with monocular vision, 2015 IEEE International Conference on Robotics and Automation (ICRA), pp. 1633-1639, (2015)
  • [6] Passani M., Yebes J.J., Bergasa L.M., Fast pixelwise road inference based on Uniformly Reweighted Belief Propagation, 2015 IEEE Intelligent Vehicles Symposium (IV), pp. 519-524, (2015)
  • [7] Ladicky L., Sturgess P., Russell C., Et al., Joint optimization for object class segmentation and dense stereo reconstruction, International Journal of Computer Vision, 100, 2, pp. 122-133, (2012)
  • [8] Sturgess P., Alahari K., Ladicky L., Et al., Combining appearance and structure from motion features for road scene understanding, BMVC 2012-23rd British Machine Vision Conference, pp. 1-10, (2012)
  • [9] Kuhnl T., Kummert F., Ffitach J., Monocular road segmentation using slow feature analysis, 2011 IEEE Conference on Intelligent Vehicles Symposium (IV), pp. 800-806, (2011)
  • [10] Xiao L., Dai B., Wu T., Et al., Unstructured road segmentation method based on dictionary learning and sparse representation, Journal of Jilin University (Engineering and Technology Edition), 43, pp. 384-388, (2013)
12