Jean Marc Lavest

Simultaneous object pose and velocity computation using a single view from a rolling shutter camera

An original concept for computing instantaneous 3D pose and 3D velocity of fast moving objects using a single view is proposed, implemented and validated. It takes advantage of the image deformations induced by rolling shutter in CMOS image sensors. First of all, after analysing the rolling shutter phenomenon, we introduce an original model of the image formation when using such a camera, based on a general model of moving rigid sets of 3D points. Using 2D-3D point correspondences, we derive two complementary methods, compensating for the rolling shutter deformations to deliver an accurate 3D pose and exploiting them to also estimate the full 3D velocity. The first solution is a general one based on non-linear optimization and bundle adjustment, usable for any object, while the second one is a closed-form linear solution valid for planar objects. The resulting algorithms enable us to transform a CMOS low cost and low power camera into an innovative and powerful velocity sensor. Finally, experimental results with real data confirm the relevance and accuracy of the approach.

Exploiting Rolling Shutter Distortions for Simultaneous Object Pose and Velocity Computation Using a Single View

An original method for computing instantaneous 3D pose and velocity of fast moving objects using a single view is presented. It exploits image deformations induced by rolling shutter in CMOS image sensors. First of all, a general perspective projection model of a moving 3D point is presented. A solution for the pose and velocity recovery problem is then described. The method is based on bundle adjustment and uses point correspondences. The resulting algorithm enables to transform a CMOS low cost and low power camera into an original velocity sensor. Finally, experimental results with real data confirm the relevance of the approach.

Self-calibration of Scheimpflug cameras: an easy protocol

The problem of the self-calibration of Scheimpflug cameras is addressed. The approach aims to calibrate by handpositioning a roughly-known calibration pattern, thanks to a bundle adjustment technique. The purpose of this paper is to design an easy protocol accordingly. A simple motion of the calibration target is introduced and a minimal number of recordings are obtained from the simulated data. The precision required for the calibration pattern is also discussed. Finally, the suggested protocol is tested in real conditions, especially in the stereo particle image velocimetry configuration.

Camera self-calibration in scheimpflug condition for air flow investigation

Camera calibration is a necessary preliminary step to ensure high quality measurement in a section of a flow. The bundle adjustment technique is based on redundancy resulting from multi-viewing the calibration target. In this paper we proposed a well-adapted protocol for self-calibrating a Scheimpflug model. Such a protocol allows calibration by hand positioning a laser printed target. Experimental results indicate that with only 5 views we can accurately calibrate a Scheimpflug camera in Stereoscopic Particle Image Velocimetry condition.

Detection and tracking vehicles using a zoom camera over a pan-and-tilt unit

The concern about safety in mobile vehicles and means of transport normally used by people has increased considerably in last decades. Different sensorial systems have been developed and improved in order to save lives, inside vehicles and outside them. In this way, imaging and vision systems can be considered to increase the visibility and detect, fix or mobile vehicles/obstacles from a moving vehicle or static/strategic position. There are different approaches to capture an image sequence from a camera and analyze them. In this paper, zoom cameras are used to improve the safety of transport systems. Having calibrated the zoom camera the goal is make measurements from far away obstacles and mobiles, increasing the precision and reliability of conventional visual systems made up of cameras with fixed lenses, fixed resolution. Besides, the zoom camera is mounted over a pan-tilt unit that gives the device the possibility to focus on different region of interest. The whole imaging system is also calibrated to reference its measurements to the global reference system.

Multiview calibration of an active vision head

In this paper a method to automatically calibrate a pan/tilt head using an imaging vision system is shown. The vision sub-system uses two cameras. Both cameras must be calibrated before. A pattern to perform the automated calibration has been constructed. The camera calibration, using that pattern, consists of two stages: capture the data online and calibrate offline in few seconds. The first task has been subdivided in another two: detection and tracking the pattern. To obtain the value of the parameters that govern the behaviour of the system, the 3D positions of the pattern points are matched in both cameras. The problem becomes a problem of minimising the position error of pattern points given by the active vision head to the real ones given by the external camera. Check and test the calibration of the system is straightforward viewing the residual errors obtained by the minimisation algorithm. Thus, an automated calibration of the active vision head based on a multiview approach has been performed.