Spatial ray features for real-time ego-lane extraction

In order to support driver assistance systems in unconstrained environments, we propose to extend local appearance-based road classification with a spatial feature generation and classification. Therefore, a hierarchical approach consisting of multiple low level base classifiers, the novel spatial feature generation, as well as a final road terrain classification, is used. The system perceives a variety of local properties of the environment by means of base classifiers operating on patches extracted from monocular camera images, each represented in a metric confidence map. The core of the proposed approach is the computation of spatial ray features (SPRAY) from these confidence maps. With this, the road-terrain classifier can decide based on local visual properties and their spatial layout in the scene. In order to show the feasibility of the approach, the extraction and evaluation of the metric ego-lane driving corridor on an inner city stream is demonstrated. This is a challenging task because on a local appearance level, ego-lane is not distinguishable from other asphalt parts on the road. However, by incorporating the proposed SPRAY features the distinction is possible without requiring an explicit lane model. Due to the parallel structure of this bottom-up approach, the implemented system operates in real-time with approximately 25 Hz on a GPU.

[1]  Yoav Freund,et al.  A decision-theoretic generalization of on-line learning and an application to boosting , 1995, EuroCOLT.

[2]  Yoav Freund,et al.  A decision-theoretic generalization of on-line learning and an application to boosting , 1997, EuroCOLT.

[3]  J. Little,et al.  Inverse perspective mapping simplifies optical flow computation and obstacle detection , 2004, Biological Cybernetics.

[4]  Guoying Zhang,et al.  A Feature Selection Algorithm Based on Boosting for Road Detection , 2008, 2008 Fifth International Conference on Fuzzy Systems and Knowledge Discovery.

[5]  Pierre Charbonnier,et al.  Evaluation of Road Marking Feature Extraction , 2008, 2008 11th International IEEE Conference on Intelligent Transportation Systems.

[6]  Bernt Schiele,et al.  A Dynamic Conditional Random Field Model for Joint Labeling of Object and Scene Classes , 2008, ECCV.

[7]  Vincent Lepetit,et al.  Fast Ray features for learning irregular shapes , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[8]  Jannik Fritsch,et al.  Adaptive multi-cue fusion for robust detection of unmarked inner-city streets , 2009, 2009 IEEE Intelligent Vehicles Symposium.

[9]  Klaus C. J. Dietmayer,et al.  Road course estimation in occupancy grids , 2010, 2010 IEEE Intelligent Vehicles Symposium.

[10]  Stefan Lüke,et al.  Map based road boundary estimation , 2010, 2010 IEEE Intelligent Vehicles Symposium.

[11]  Theo Gevers,et al.  3D Scene priors for road detection , 2010, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[12]  Takashi Naito,et al.  Multiband Image Segmentation and Object Recognition for Understanding Road Scenes , 2011, IEEE Transactions on Intelligent Transportation Systems.

[13]  Franz Kummert,et al.  Monocular road segmentation using slow feature analysis , 2011, 2011 IEEE Intelligent Vehicles Symposium (IV).

[14]  Wolfgang Förstner,et al.  A temporal filter approach for detection and reconstruction of curbs and road surfaces based on Conditional Random Fields , 2011, 2011 IEEE Intelligent Vehicles Symposium (IV).

[15]  Sergiu Nedevschi,et al.  New results in stereovision based lane tracking , 2011, 2011 IEEE Intelligent Vehicles Symposium (IV).

[16]  Thomas Gumpp,et al.  Lane confidence fusion for visual occupancy estimation , 2011, 2011 IEEE Intelligent Vehicles Symposium (IV).

[17]  Tao Wu,et al.  A practical system for road marking detection and recognition , 2012, 2012 IEEE Intelligent Vehicles Symposium.

[18]  Andrew W. Fitzgibbon,et al.  Real-time human pose recognition in parts from single depth images , 2011, CVPR 2011.