scholarly journals Fluid-inspired field representation for risk assessment in road scenes

2020 ◽  
Vol 6 (4) ◽  
pp. 401-415
Author(s):  
Xuanpeng Li ◽  
Lifeng Zhu ◽  
Qifan Xue ◽  
Dong Wang ◽  
Yongjie Jessica Zhang

AbstractPrediction of the likely evolution of traffic scenes is a challenging task because of high uncertainties from sensing technology and the dynamic environment. It leads to failure of motion planning for intelligent agents like autonomous vehicles. In this paper, we propose a fluid-inspired model to estimate collision risk in road scenes. Multi-object states are detected and tracked, and then a stable fluid model is adopted to construct the risk field. Objects’ state spaces are used as the boundary conditions in the simulation of advection and diffusion processes. We have evaluated our approach on the public KITTI dataset; our model can provide predictions in the cases of misdetection and tracking error caused by occlusion. It proves a promising approach for collision risk assessment in road scenes.

Author(s):  
Xuanpeng Li ◽  
Lifeng Zhu

Prediction of the likely evolution in the traffic scenes is a challenging task because of high uncertainty of sensing technology and dynamic environment. It leads to failure of planning for intelligent agents like autonomous vehicles. In this paper, we propose a fluid-based physical model to present the influence of surrounding object's motion on driving safety. In our pipeline, the input sensor could be LiDAR, camera, or multi-modal data. We use a Kalman filter to estimate the state space of each detected object, and adopt the properties of stable fluid to build a riskmap based on the density field. The noisy state space are then modeled as the boundary conditions in the simulation of advection and diffusion process. We test our approach on the public KITTI dataset and find this model could handle the short-term prediction in case of misdetection and tracking failure caused by object occlusion, which shows promising in collision risk assessment on road scenes.


2021 ◽  
Author(s):  
Weimin Qi ◽  
Qinbo Sun ◽  
Chongfeng Liu ◽  
Xiaoqiang Ji ◽  
Zhongzhong Cao ◽  
...  

Robotica ◽  
2012 ◽  
Vol 31 (4) ◽  
pp. 525-537 ◽  
Author(s):  
F. Belkhouche ◽  
B. Bendjilali

SUMMARYThis paper introduces a probabilistic model for collision risk assessment between moving vehicles. The uncertainties in the states and the geometric variables obtained from the sensory system are characterized by probability density functions. Given the states and their uncertainties, the goal is to determine the probability of collision in a dynamic environment. Two approaches are discussed: (1) The virtual configuration space (VCS), and (2) the rates of change of the visibility angles. The VCS is a transformation of observer that reduces collision detection with a moving object to collision detection with a stationary object. This approach allows to create simple geometric collision cones. Error propagation models are used to solve the problem when going from the VCS to the configuration space. The second approach derives the collision conditions in terms of the rate of change of the limit visibility angles. The probability of collision is then calculated. A comparison between the two methods is carried out. Results are illustrated using simulation, including Monte Carlo simulation.


2019 ◽  
Vol 72 (06) ◽  
pp. 1449-1468 ◽  
Author(s):  
Weibin Zhang ◽  
Xinyu Feng ◽  
Yong Qi ◽  
Feng Shu ◽  
Yijin Zhang ◽  
...  

The absence of a regional, open water vessel collision risk assessment system endangers maritime traffic and hampers safety management. Most recent studies have analysed the risk of collision for a pair of vessels and propose micro-level risk models. This study proposes a new method that combines density complexity and a multi-vessel collision risk operator for assessing regional vessel collision risk. This regional model considers spatial and temporal features of vessel trajectory in an open water area and assesses multi-vessel near-miss collision risk through danger probabilities and possible consequences of collision risks via four types of possible relative striking positions. Finally, the clustering method of multi-vessel encountering risk, based on the proposed model, is used to identify high-risk collision areas, which allow reliable and accurate analysis to aid implementation of safety measures.


Sign in / Sign up

Export Citation Format

Share Document