Bayesian curved lane estimation for autonomous driving
- PDF / 3,944,150 Bytes
- 11 Pages / 595.276 x 790.866 pts Page_size
- 98 Downloads / 294 Views
ORIGINAL RESEARCH
Bayesian curved lane estimation for autonomous driving Mohamed Fakhfakh1 · Lotfi Chaari1,2 · Nizar Fakhfakh3 Received: 9 April 2019 / Accepted: 3 January 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract Several pieces of research during the last decade in intelligent perception are focused on the development of algorithms allowing vehicles to move efficiently in complex environments. Most of existing approaches suffer from either processing time which do not meet real-time requirements, or inefficient in real complex environment, which also does not meet the full availability constraint of such a critical function. To improve the existing solutions, an algorithm based on curved lane detection by using a Bayesian framework for the estimation of multi-hyperbola parameters is proposed to detect curved lane under challenging conditions. The general idea is to divide a captured image into several parts. The trajectory is modeled by a hyperbola over each part, whose parameters are estimated using the proposed hierarchical Bayesian model. Compared to the existing works in the state of the art, experimental results prove that our approach is more efficient and more precise in road marking detection. Keywords Autonomous driving · Embedded camera · Road marking · Multi-hyperbola · Bayesian framework
1 Introduction Autonomous driving is one of the main concerns which allows a step by step evolution of Artificial Intelligence. Indeed, it is important to ensure the proper functioning of autonomous vehicle mechanisms by making appropriate and accurate decisions in real time by identifying and distinguishing the different objects of the road. In this direction, several studies have been carried out on autonomous driving in order to improve road security. The methods of obtaining intelligent vehicles systems are solely based on artificial intelligence, by trying reproducing human intelligence, and more precisely Machine Learning/ Deep leaning approaches. The main goal of research and development of smart vehicles is to reduce accident rates,
* Mohamed Fakhfakh [email protected] Lotfi Chaari lotfi.chaari@toulouse‑inp.fr Nizar Fakhfakh [email protected] 1
MIRACL Laboratory, University of Sfax, Sfax, Tunisia
2
IRIT‑ENSEEIHT, University of Toulouse, Toulouse, France
3
NAVYA, Paris, France
as well as improve the efficiency of traffic use by detecting different dangerous situations. As regards the data, the state of the art studies generally rely on embedded equipments on the autonomous vehicles such as cameras [mono (Nguyen et al. 2015), stereo (Song et al. 2018) or multi-camera (Ieng et al. 2005)] or other sensors like Odometer (Santana et al. 2013), IMU (Anwary et al. 2018), Differential GPS (Chengping et al. 2014 or Kim and Park 2017b). Embedded processors allow running algorithms in order to move successfully from a starting point to a predefined destination. The main difficulties to perform this task generally lie in the possible presence of
Data Loading...
