Decision-free true positive estimation with grid maps for multi-object tracking

Nuß, Dominik; Wilking, Benjamin; Wiest, Jürgen; Deusch, Hendrik; Reuter, Stephan; Dietmayer, Klaus

doi:10.1109/itsc.2013.6728206

Cited by 15 publications

(13 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is possible with the Dempster-Shafer theory of evidence [12], [13], [14]. For the application of the Dempster-Shafer theory of evidence on grid maps and a general introduction, see also [15] and [16].…”

Section: A Representing Dempster-shafer Occupancy Mass With Particlesmentioning

confidence: 99%

See 1 more Smart Citation

Fusion of laser and radar sensor data with a sequential Monte Carlo Bayesian occupancy filter

Nuß

Yuan

Krehl

et al. 2015

2015 IEEE Intelligent Vehicles Symposium (IV)

Self Cite

View full text Add to dashboard Cite

Occupancy grid mapping is a well-known environment perception approach. A grid map divides the environment into cells and estimates the occupancy probability of each cell based on sensor measurements. An important extension is the Bayesian occupancy filter (BOF), which additionally estimates the dynamic state of grid cells and allows modeling changing environments. In recent years, the BOF attracted more and more attention, especially sequential Monte Carlo implementations (SMC-BOF), requiring less computational costs. An advantage compared to classical object tracking approaches is the objectfree representation of arbitrarily shaped obstacles and freespace areas. Unfortunately, publications about BOF based on laser measurements report that grid cells representing big, contiguous, stationary obstacles are often mistaken as moving with the velocity of the ego vehicle (ghost movements). This paper presents a method to fuse laser and radar measurement data with the SMC-BOF. It shows that the doppler information of radar measurements significantly improves the dynamic estimation of the grid map, reduces ghost movements, and in general leads to a faster convergence of the dynamic estimation.

show abstract

Section: A Representing Dempster-shafer Occupancy Mass With Particlesmentioning

confidence: 99%

“…In order to calculate the posterior occupancy BBA m(o c k+1|k+1 ), we combine the observation-based BBA and the predicted BBA individually for each grid cell c using the Dempster-Shafer combination rule (see e.g. [15]):…”

Section: B Using Occupancy Measurement Gridsmentioning

confidence: 99%

Fusion of laser and radar sensor data with a sequential Monte Carlo Bayesian occupancy filter

Nuß

Yuan

Krehl

et al. 2015

2015 IEEE Intelligent Vehicles Symposium (IV)

Self Cite

View full text Add to dashboard Cite

show abstract

“…In addition, the Dempster-Shafer theory of evidence (Shafer 1976;Zou et al 2000;Nuss et al 2013;Effertz 2008) is increasingly establishing itself in the field of grid mapping. It allows differentiation between uncertainty due to probability theory and uncertainty due to lack of information.…”

Section: Use Of Lidar or Radar Datamentioning

confidence: 99%

Representation of Fused Environment Data

Dietmayer

Reuter

Nuß

2015

Handbook of Driver Assistance Systems

Self Cite

View full text Add to dashboard Cite

The requirements for a vehicle environment representation increase with the complexity of advanced driver assistance systems and automatic driving. The ability of the current traffic situation to interpret and predict is essential for being able to automatically derive reasonable decisions. As a consequence, a state of the art vehicle environment representation has to incorporate all relevant dynamic objects as well as static obstacles and context information. While dynamic objects are typically described by an objectbased representation using state variables, static obstacles as well as free space area are commonly modeled using grid-based methods. This chapter gives an introduction into both of these concepts.The chapter is organized as follows: First, the difference between function-oriented and modular fusion architectures is discussed. Afterwards, the joint integrated probabilistic data association (JIPDA) filter is introduced, which is one method to realize an object-based environment model incorporating both state and existence uncertainties. Further, the representation of static obstacles with occupancy grids is described in detail and the incorporation of measurements of different sensor types is illustrated. Finally, several hybrid environment representations are introduced and an example for a strictly modular architecture, the hierarchical modular environment perception, is presented. Requirements for Vehicle Environment RepresentationsA vehicle environment representation, often also referred to as a vehicle environment model, is understood to be a dynamic data structure in which all relevant objects and infrastructure elements in the proximity of the ego vehicle are contained. All elements have to be represented in a common reference system as accurately as possible with regard to position and time. The detection and the temporal tracking of the objects and infrastructure elements are performed continuously using on-board sensors such as cameras and radar (see chapters "▶ Automotive Radar", "▶ Automotive Lidar", "▶ 3-D-Imaging for ADAS", "▶ Automotive Camera (Hardware)" and "▶ Fundamentals of Machine Vision"). In the future increasingly more information from high-precision, attributed digital maps and, if applicable, external information based on Car2x communication will be available and can be incorporated in the data fusion. Figure 1 shows examples of common elements contained in a vehicle environment representation.The objects and structure elements which are relevant for a vehicle environment representation depend to a large extent on the functions that will be implementing it. For example, a blind spot assist requires only the information on whether there are currently objects in the rear or side area of the vehicle, the type of object is immaterial. More complex assistance systems, for example automatic emergency steer assist all the way through to automatic driving, require more extensive perceptiveness and information. In these cases, the distances, speeds, and dimensions of all traff...

show abstract

“…Thus, we use the combination rule defined in (9) to fuse an accumulated grid of the stereo camera. In order to fuse a global grid from the accumulated grids of the laser scanner and the stereo camera we also use the combination rule defined in (9). However, we apply the modification that the conflicting mass between m(F ) and any other hypothesis is assigned to m acc (F ).…”

Section: E Grid Fusion Algorithmmentioning

confidence: 99%

“…The fusion approach was used to identify the shape of the road to navigate an autonomous car. Nuss et al used a height threshold to prevent that cells that are occupied with small obstacles are falsely classified as free space [9]. The method proposed in this work is based on grid mapping using Dempster-Shafer's theory (DST) of evidence [10] [11].…”

Section: Introductionmentioning

confidence: 99%

2-D Evidential Grid Mapping with Narrow Vertical Field of View Sensors Using Multiple Hypotheses and Spatial Neighborhoods

Seeger

Manz

Matters

et al. 2015

2015 IEEE 18th International Conference on Intelligent Transportation Systems

View full text Add to dashboard Cite

Environment sensors with a narrow vertical field of view often fail to detect obstacles with a small vertical extent from close range. Because an inverse beam sensor model infers free space when there was no measurement, those obstacles are deleted from an occupancy grid eventhough they have been observed in past measurements. This is extremely critical if the car is driving autonomously. Our approach explicitly models these errors using multiple hypotheses in an evidential grid mapping framework that neither needs a classification nor a height of obstacles. Furthermore, we extend the grid mapping framework, that usually assumes mutually independent cells, by including information from neighboring cells. The evaluation in a freeway and a challenging scenario with a boom barrier shows that our method is superior to evidential grid mapping with an inverse beam sensor model.

show abstract

Decision-free true positive estimation with grid maps for multi-object tracking

Cited by 15 publications

References 15 publications

Fusion of laser and radar sensor data with a sequential Monte Carlo Bayesian occupancy filter

Fusion of laser and radar sensor data with a sequential Monte Carlo Bayesian occupancy filter

Representation of Fused Environment Data

2-D Evidential Grid Mapping with Narrow Vertical Field of View Sensors Using Multiple Hypotheses and Spatial Neighborhoods

Contact Info

Product

Resources

About