Autonomous driving is based on innovative technologies that have to ensure vehicles are driven safely. LiDARs one of the reference sensors for obstacle detection. However, this technology affected by adverse weather conditions, especially fog. Different wavelengths investigated meet challenge (905 nm vs. 1550 nm). The influence wavelength light transmission in fog then examined and results reported. A theoretical approach calculating extinction coefficient different presented comparison...

Computer vision systems are increasingly present in road environments. These have not been evaluated adverse weather conditions, particularly rain. The objective of this article is to develop tools validate these computer such conditions. This study begins by setting up a digital rain image simulator based on set physical laws. In addition the simple visual effect, it makes possible obtain images displaying reality. was then validated against data acquired Cerema R&D Fog and Rain platform....

Object detection is recognized as one of the most critical research areas for perception self-driving cars. Current vision systems combine visible imaging, LIDAR, and/or RADAR technology, allowing vehicle’s surroundings. However, harsh weather conditions mitigate performances these systems. Under circumstances, thermal imaging becomes complementary solution to current not only because it makes possible detect and recognize environment in extreme conditions, but also images are compatible...

Improving the reliability of automotive perceptive sensors in degraded weather conditions, including fog, is an important issue for road safety and development automated driving. Cerema has designed PAVIN platform reproducing fog rain conditions to evaluate optical sensor performance under these conditions. In order increase variety scenarios technologies test, use digital simulation becomes a major asset. The purpose this paper revive debate around realism various models underlying...

Fog is one of major challenges for transportation systems. The automation the latter based on perception sensors that can be disrupted by atmospheric conditions. As fog conditions are random and non-reproducible in nature, Cerema has designed a platform to generate rain demand. Two types artificial with different droplet size distributions generated: they correspond radiation fogs small medium droplets. This study presents an original method classifying these descriptive quantitative way. It...

This paper presents a novel calibration algorithm for plenoptic cameras, especially the multi-focus configuration, where several types of micro-lenses are used, using raw images only. Current methods rely on simplified projection models, use features from reconstructed images, or require separated calibrations each type micro-lens. In same part scene will demonstrate different amounts blur according to micro-lens focal length. Usually, only micro-images with smallest amount used. order...

In road environments, real-time knowledge of local weather conditions is an essential prerequisite for addressing the twin challenges enhancing safety and avoiding congestions. Currently, main means quantifying along a network requires installation meteorological stations. Such stations are costly must be maintained; however, large numbers cameras already installed on roadside. A new artificial intelligence method that uses traffic convolution neural to detect has, therefore, been proposed....

A wide range of sensors is being developed for autonomous vehicles in order to ensure safe driving conditions by detecting any relevant objects the road environment. Sensors, cameras, Lidars and radars, cover a wavelength, ranging from visible, Near Infrared, Short Wave Infrared beyond high frequency radars. For Lidar technology, single wavelength scans To obtain most efficient detection, choice must be optimized. Knowledge spectral reflectance at chosen key point. An original application...

This work aims at introducing modelling, theoretical and numerical studies related to a new downscaling technique applied computational fluid dynamics. Our method consists in building local model, forced by large scale information computed thanks classical weather predictor. The compatible with the Navier-Stokes equations, is used for small computation (downscaling) of considered fluid. It inspired Pope's works on turbulence, so-called Langevin system stochastic differential equations. We...

Hydraulic snow melting system using pavement structure has been employed for some critical roads in order to prevent ice and formation. This paper describes a 2D hydrothermal numerical model of hydraulic that circulates fluid through bonding porous asphalt layer. A set boundary conditions have defined, heat mass transfer developed. can predict temperature field surface when the Model development is described validated this data collected from full-scale system.

This paper presents a novel calibration algorithm for Multi-Focus Plenoptic Cameras (MFPCs) using raw images only. The design of such cameras is usually complex and relies on precise placement optic elements. Several procedures have been proposed to retrieve the camera parameters but relying simplified models, reconstructed extract features, or multiple calibrations when several types micro-lens are used. Considering blur information, we propose new Blur Aware (BAP) feature. It first...

A Lagrangian stochastic model is introduced in order to describe the local behavior of wind. Based on some MM5 data, and thanks particles driven by differential equations, we propose a numerical method allowing improve simulations at small scales, without requiring too much additional computational cost.

Nowadays, many pedestrians are victims of road accidents. Several artificial vision solutions, based on pedestrian detection, have therefore been developed to assist drivers and reduce the accident rate. But most proposed databases make it possible test detection only in favorable conditions. The main goal this research is provide a learning testing environment for development detectors able function under all weather conditions by day even night. This paper presents new database, called...