The article deals with the analysis and interpretation of dynamic scenes typical of urban driving. The key objective is to assess risks of collision for the ego-vehicle. We describe our concept and methods, which we have integrated and tested on our experimental platform on a Lexus car and a driving simulator. The on-board sensors deliver visual, telemetric and inertial data for environment monitoring. The sensor fusion uses our Bayesian Occupancy Filter for a spatio-temporal grid representation of the traffic scene. The underlying probabilistic approach is capable of dealing with uncertainties when modeling the environment as well as detecting and tracking dynamic objects. The collision risks are estimated as stochastic variables and are predicted for a short period ahead with the use of Hidden Markov Models and Gaussian processes. The software implementation takes advantage of our methods, which allow for parallel computation. Our tests have proven the relevance and feasibility of our approach for improving the safety of car driving.
Abstract-Modeling and monitoring dynamic environments is a complex task but is crucial in the field of intelligent vehicle. A traditional way of addressing these issues is the modeling of moving objects, through Detection And Tracking of Moving Objects (DATMO) methods. An alternative to a classic object model framework is the occupancy grid filtering domain. Instead of segmenting the scene into objects and track them, the environment is represented as a regular grid of occupancy, in which each cell is tracked at a sub-object level. The Bayesian Occupancy Filter [1] is a generic occupancy grid framework which predicts the spread of spatial occupancy by estimating cell velocity distributions. However its velocity model, corresponding to a transition histogram per cell, leads to huge data management which in practice makes it hardly compatible to severe computational and hardware constraints, like in many embedded systems. In this paper, we present a new representation for the BOF, describing the environment through a mix of static and dynamic occupancy. This differentiation enables the use of a model adapted to the considered nature: static occupancy is described in a classic occupancy grid, while dynamic occupancy is modeled by a set of moving particles. Both static and dynamic parts are jointly generated and evaluated, their distribution over the cells being adjusted. This approach leads to a more compact model and to drastically improve the accuracy of the results, in particular in term of velocities. Experimental results show that the number of values required to model the velocities have been reduced from a typical 900 per cell (for a 30x30 neighborhood) to less than 2 per cell in average. The massive data compression allows to plan dedicated embedded devices.
Proper modeling of dynamic environments is a core task in the field of intelligent vehicles. The most common approaches involve the modeling of moving objects, through Detection And Tracking of Moving Objects (DATMO) methods. An alternative to a classic object model framework is the occupancy grid filtering domain. Instead of segmenting the scene into objects and track them, the environment is represented as a regular grid of occupancy, in which spatial occupancy is tracked at a sub-object level. In this paper, we present the Conditional Monte Carlo Dense Occupancy Tracker, a generic spatial occupancy tracker, which infers dynamics of the scene through an hybrid representation of the environment, consisting of static occupancy, dynamic occupancy, empty spaces and unknown areas. This differentiation enables the use of state specific models (classic occupancy grid for motion-less components, set of moving particles for dynamic occupancy) as well as proper confidence estimation and management of data-less areas. The approach leads to a compact model that drastically improves the accuracy of the results and the global efficiency in comparison to previous methods.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.