Stéphane Mousset

Vehicle detection by means of stereo vision-based obstacles features extraction and monocular pattern analysis

This paper presents a stereo vision system for the detection and distance computation of a preceding vehicle. It is divided in two major steps. Initially, a stereo vision-based algorithm is used to extract relevant three-dimensional (3-D) features in the scene, these features are investigated further in order to select the ones that belong to vertical objects only and not to the road or background. These 3-D vertical features are then used as a starting point for preceding vehicle detection; by using a symmetry operator, a match against a simplified model of a rear vehicles shape is performed using a monocular vision-based approach that allows the identification of a preceding vehicle. In addition, using the 3-D information previously extracted, an accurate distance computation is performed.

A cooperative approach to vision-based vehicle detection

In this paper two different vision based systems for vehicle detection are described and their integration discussed. The first approach is based on the use of a specific model for vehicles and mostly relies on monocular vision. Conversely, the second system is based on the use of stereo vision and allows to refine the coarse results obtained by the former. A preliminary integration of the two systems has been tested on the ARGO experimental vehicle and some remarks about reliability and robustness are also included.

Stereo vision-based feature extraction for vehicle detection

This paper presents a stereo vision system for vehicle detection. It has been conceived as the integration of two different subsystems. Initially a stereo vision based system is used to recover the most relevant 3D features in the scene; due to the algorithms generality, all the vertical features are extracted as potentially belonging to a vehicle in front of the vision system. This list of significant patterns is fed to a second subsystem based on monocular vision; it processes the list computing a match with a general model of a vehicle based on symmetry and shape, thus allowing the identification of the sole characteristics belonging to a vehicle. The system presented in this work derives from the integration of the research work developed by the University of Parma (Italy) and I.N.S.A. of Rouen (France). The two subsystems have been integrated into the GOLD software and are currently under testing using the ARGO experimental vehicle.

An application of V2V communications : Cooperation of vehicles for a better car tracking using GPS and vision systems

The advanced driver assistance systems (ADAS systems) based on cooperation between vehicles can offer serious perspectives to the road security. The inter-vehicle cooperation is made possible thanks to the revolution in the wireless mobile ad hoc network. This paper presents an application of the V2V communications. The car tracking using the GPS receivers is not always ideal in urban areas. To resolve this problem, our original approach is based on using data issued from vision systems when the GPS signal is not available or has a poor quality due to multi-tracks or bad satellite visibility. The method relies on the particle filter for the fusion of the GPS data and the vision data that will be collected from the loading system in the vehicles. Multiple results from different scenarios experimented on our fleet of communicating vehicles carried out in real conditions prove the feasibility of this approach.

Cooperation of passive vision systems in detection and tracking of pedestrians

This work presents a cooperative approach for detecting and tracking pedestrians in an urban environment. Its originality lies in the cooperation of two vision systems. A monocular vision system retrieves feature elements and these elements are visualized. However, false detection can occur due to objects whose outline is similar to that of a pedestrian. This problem is solved by the introduction of an auto-adaptive stereovision algorithm that recovers all the vertical 3D segments of the scene. This cooperation supplies a fast and robust method for detecting pedestrian presence. Then, it allows for pedestrian tracking through multiple images.

Temporal consistent fast stereo matching for advanced driver assistance systems (ADAS)

In this paper, we present a new fast method for matching stereo images acquired by a stereo sensor embedded in a moving vehicle. The method consists in exploiting the matching results obtained in one stereo pair (frame) for computing the disparity map of the following stereo pair. This can be achieved by finding a temporal relationship, which we named association, between consecutive frames. The disparity range of the current frame is deduced from the disparity map of the preceding frame and the association between the two frames. Dynamic programming technique is considered for matching the image features. The proposed approach is tested on virtual and real stereo image sequences and the results are satisfactory. The method is fast and able to provide about 20 millions disparity maps per second on a HP Pavilion dv6700 2.1GHZ.

A new stereo matching approach for real-time road obstacle detection for situations with deteriorated visibility

This paper presents a fast stereo matching approach for road obstacle detection under foggy weather conditions. The stereo matching process can be treated as the problem of finding an optimal path on a 2D search plane. To obtain this path, we propose a new cost function. This last is derived from the variance values of the intensities on the right hand sides of the matched declivities. The matching process is executed independently for each scanline. In order to reduce the false matches and speed up the matching process, we propose to exploit the relationship between successive stereo images. So, the disparity map computed for one stereo pair will be used to find the disparity range for the next stereo pair. The disparity range is deduced for each scanline. The proposed approach has been tested on synthesized and real images under foggy weather conditions. The new method gives satisfactory results.

An application of the Sequential Monte Carlo to increase the accuracy of travel time estimation in urban areas

This paper presents an application of the Sequential Monte Carlo that will help to increase the accuracy of travel time estimations in our historical data. Our estimation filter is based on the Monte Carlo Method and was modeled in such a way as to be applicable to our new kind of data in order to estimate travel time per section of road. We took into consideration the delay time while changing the sections to symbolize the delay due to traffic lights or crossroads. We worked on an urban zone of Rouen, a French city, to evaluate our application. In this application, information is collected from a specific GPS system that warns drivers of the location of both fixed and mobile speed radars. Unlike the classical GPS system, this system is characterized by the data flow frequency where the GPS data is received from the probe vehicles at one minute intervals. After receiving the data we apply the map matching method in order to correct the GPS errors. Also, our geo-referencing system has special features; each road or section of road is formed by nodes and segments, and the intersection between each section is called a PUMAS points. The PUMAS Points are GPS coordinate points on a digital map which can be propagated or moved without cost, providing total flexibility to mesh a city or rural area. Over all the performance of the filter estimator is around 85% if we set our threshold at 50%.

Visual odometry with unsynchronized multi-cameras setup for intelligent vehicle application

This paper presents a visual odometry with metric scale estimation of a multi-camera system in challenging un-synchronized setup. The intended application is in the field of intelligent vehicles. We propose a new algorithm named “triangle-based” method. The proposed algorithm employs the information from both extrinsic and intrinsic parameters of calibrated cameras. We assume that the trajectory between two consecutive frames of a camera is a linear segment (straight trajectory). The relative camera poses are estimated via classical Structure-from-Motion. Then, the scale factors are computed by imposing the known extrinsic parameters and the linearity assumption. We verify the validity of our method both in simulated and real conditions. For the real world, the motion trajectory estimated for image sequence of two cameras from KITTI dataset is compared against the GPS/INS ground truth.

Evaluating new prototyping framework based on combining embedded sensors data and ns2 simulations

ADAS systems become increasingly integrated and provide more comprehensive driver support using information coming from surrounding environment and sensors for the modeling of the global environment. The evolution of technologies over the coming years will involve applications usingV2V and V2I communications. The optimization of these applications tosupport driving systems is not completed because the communications between vehicles are not well controlled. This paper aims at encouraging the cooperation between two software systems. An integrated framework for prototyping new communicating ”ADAS” systems is presented. This framework will achieve the communication and the data merging from different systems. It consists of two components, a Real-Time Multi-sensors Advanced Prototyping (RTMaps) that realizes data acquisition from different sources, and a network simulator (ns2) that simulates wireless packet transmission. We willassessthe impact ofcommunication in this new framework of ADAS by combining embedded sensors data (GPS-vision) and V2V simulations towards collision avoidance application. Multiple results from different scenarios are validated to prove the feasibility and the performance efficiency of real-time multi sensors. We assess the whole system under different traffic scenarios and different routing algorithms.