Search

Much work is currently devoted to increasing the reliability, completeness and precision of the data used by driving assistance systems, particularly in urban environments. Urban environments represent a particular challenge for the task of perception, since they are complex, dynamic and completely variable. This article examines a multi-modal perception approach for enhancing vehicle localization and the tracking of dynamic objects in a world-centric map. 3D ego-localization is achieved by merging stereo vision perception data and proprioceptive information from vehicle sensors. Mobile objects are detected using a multi-layer lidar that is simultaneously used to identify a zone of interest to reduce the complexity of the perception process. Object localization and tracking is then performed in a fixed frame which simplifies analysis and understanding of the scene. Finally, tracked objects are confirmed by vision using 3D dense reconstruction in focused regions of interest. Only confirmed objects can generate an alarm or an action on the vehicle. This is crucial to reduce false alarms that affect the trust that the driver places in the driving assistance system. Synchronization issues between the sensing modalities are solved using predictive filtering. Real experimental results are reported so that the performance of the multi-modal system may be evaluated. [ABSTRACT FROM AUTHOR]

A Robust Set-Inversion via Interval Analysis method in a bounded-error framework is used to compute three-dimensional location zones in real time, at a given confidence level. This approach differs significantly from the usual Gaussian error model paradigm, since the satellite positions and the pseudorange measurements are represented by intervals encompassing the true value with a particular level of confidence. The method computes a location zone recursively, using contractions and bisections of an arbitrarily large initial location box. Such an approach can also handle an arbitrary number of erroneous measurements using a q-relaxed solver and allows the integration of geographic and cartographic information such as digital elevation models or three-dimensional maps. With enough data redundancy, inconsistent measurements can be detected and even rejected. The integrity risk of the location zone comes only from the measurement bounds settings, since the solver is guaranteed. A method for setting these bounds for a particular location zone confidence level is proposed. An experimental validation using real L1 code measurements and a digital elevation model is also reported in order to illustrate the performance of the method on real data. [ABSTRACT FROM AUTHOR]

This paper presents a loosely coupled fusion approach that merges GPS data, dead-reckoned sensors, and GIS (geographical information system) data. The GPS latency is compensated for by the DR sensors and the use of an xPPS signal. The fusion of the estimate with the map data is spatial-triggered, while the fusion with the GPS is time-triggered. We present a strategy that relies on pose tracking which is reinitialized when GPS data become incoherent. Particular attention is given to the representation of the road map and to the management of a cache memory for efficiency purposes. We report experiments carried out with our equipped car and a GIS whose characteristics are well adapted to embedded constraints. [ABSTRACT FROM AUTHOR]