In this paper, we deal with an original advanced driver assistance system (ADAS) based on the use of omnidirectional vision and an evidential fusion architecture. The panoramic perception solution permits us to address efficiently the problem of close vehicles detection but also the monitoring side traffic system. The fusion and integration of this sensorial data stream is assumed by a credibilist architecture based on the transferable belief model (TBM) of SMETS. This paradigm permits the filtering of false alarms efficiently by an optimal management of the uncertainties estimation.

This paper deals with the design of a robotic platform providing a grasping procedure to assist disabled persons. Like most of the vision-based systems, the camera used to guide the arm toward the object is embedded onto the gripper, which gives a very local view of the environment. We propose to use a catadioptric sensor in order to deal with situations where the object of interest is out of the embedded camera field of view. The mirror used has been especially designed for our robot, satisfying several constraints of the final application. The solution proposed respects one of the main matters of all systems designed for disabled persons, reduce the user strain. A single click is necessary to position the robotic arm toward the direction of the object. Clinical evaluations realized during one month in four medical centers with about 40 users confirm the validity of our system.

This article concerns road safety and driving assistance. To solve this problem, we propose a data fusion architecture based on the Dempster-Shafer theory. This multilevel approach allows the management of complementary and redundant data which come from two perception systems: an omnidirectional vision sensor and a laser telemeter. The originality of this architecture is its ability to manage and propagate uncertainties from low level data until a high level information of danger given to the driver. The first part concerns the data sensor. The second part deals with the quantification of the uncertainties of the detected vehicles, followed by a determination of situations of danger and the evaluation of their level of dangerousness with the aim of supplying the driver with an indicator of global danger around the vehicle.

Robotic researches can contribute to the restoration of some functions lost by handicapped people. The over-cost generated by the additive potentialities must be affordable and related to the value of the usual product. In most cases, autonomous functions are direct transpositions of solutions applied in industrial robotics. If we consider that in addition with cost, security is a supplementary constraint of rehabilitation robotics, an important research effort is needed to propose technological components. The aim of this paper is to introduce a functionality within the context of the following plan: "technical achievement and psychological analysis of a master/slave robot assistance for the invalid person" (HTSC: human/technology complex system). This application allows a severely handicapped person to handle at the same time a movable platform and a Manus/spl reg/ arm. In this paper, we will present an "automatic tracking" functionality we developed in order to keep the Manus/spl reg/ arm automatically close to the patient.

This paper deals with the localization and map building paradigm in an unknown indoor environment. We propose an exploration method based on the use of the sensorial data provided by an omnidirectional stereoscopic vision system. First, we link the problem of sensorial model construction with two omnidirectional images. We propose an approach based on the fusion of several criteria, which is realized according to Dempster-Shafer rules. We then deal with the matching problem of the stereo sensorial model with an environment map integrating all the previous primitive observations. We propose two matching approaches based on different selection criteria: the Hausdorff distance and the cumulated Cartesian distance. We also present our incremental map building paradigm based on the hypothesis of a non a priori knowledge. Finally, we deal with the problem of allowing a robot to localize itself and to construct concurrently a representation of its environment.

Deals with an absolute mobile robot self-localization algorithm given a two-dimensional cartography of the robot's environment. The proposed method only requires the goniometrical observations. We present an original solution to compute the robot's pose estimation by an interpretation tree (IT) search. Our approach introduces the subdivision of the robot's evolution field into rectangles. The matching between the observations and the visual landmarks is effected by judicious angular tests (heuristics for the IT). Moreover, we show that our matching method implicitly takes uncertainties and inaccuracies into account. Finally, some experimental results with real noisy omnidirectional images, provided by our omnidirectional vision sensor SYCLOP, are shown.

We present a complete localisation method for mobile robots. We deal with low level image processing, model recognition through an invariant, matching with real noisy goniometrical observations and location estimation through a triangulation technique. We show how useful invariants may be in order to solve the matching problem. Moreover, we point out how invariants are powerful to discriminate features (in images) in order to provide a quick and reliable robot location. Our localisation technique is absolute, i.e. it needs neither a (exterior) position prediction nor a resetting technique. Finally, we present experimental results with a substantial set of real noisy images that demonstrate the reliability of our localisation method.

We first present a short description of our conic sensor. Furthermore, we pay special attention to the projective invariant, represented by the cross-ratio. We end the paper by presenting a new 1D cross-ratio application that enables us to localise our mobile robot SARAH efficiently. With the method that we propose using, the camera does not need to be calibrated and this vision system, based on an omnidirectional sensor, is a major advantage for the navigation of an autonomous mobile robot.

In this paper we present a dynamic localization system which allows a mobile robot to evolve autonomously in a structured environment. Our system is based on the use of two sensors: an odometer and an omnidirectional vision system which gives a reference in connection with a set of natural beacons. Our navigation algorithm gives a reliable position estimation due to a systematic dynamic resetting. To merge the data obtained we use the extended Kalman filter. Our proposed method allows us to treat efficiently the noise problems linked to the primitive extraction, which contributes to the robustness of our system. Thus, we have developed a reliable and quick navigation system which can deals with the constraints of moving the robots in an industrial environment. We give the experimental results obtained from a mission realized in an a priori known environment.