Steven S. Beauchemin

Performance of optical flow techniques

While different optical flow techniques continue to appear, there has been a lack of quantitative evaluation of existing methods. For a common set of real and synthetic image sequences, we report the results of a number of regularly cited optical flow techniques, including instances of differential, matching, energy-based, and phase-based methods. Our comparisons are primarily empirical, and concentrate on the accuracy, reliability, and density of the velocity measurements; they show that performance can differ significantly among the techniques we implemented.

The computation of optical flow

Two-dimensional image motion is the projection of the three-dimensional motion of objects, relative to a visual sensor, onto its image plane. Sequences of time-orderedimages allow the estimation of projected two-dimensional image motion as either instantaneous image velocities or discrete image displacements. These are usually called the optical flow field or the image velocity field. Provided that optical flow is a reliable approximation to two-dimensional image motion, it may then be used to recover the three-dimensional motion of the visual sensor (to within a scale factor) and the three-dimensional surface structure (shape or relative depth) through assumptions concerning the structure of the optical flow field, the three-dimensional environment, and the motion of the sensor. Optical flow may also be used to perform motion detection, object segmentation, time-to-collision and focus of expansion calculations, motion compensated encoding, and stereo disparity measurement. We investigate the computation of optical flow in this survey: widely known methods for estimating optical flow are classified and examined by scrutinizing the hypothesis and assumptions they use. The survey concludes with a discussion of current research issues.

Detection and characterization of multiple motion points

The computation of optical flow is a well studied topic in biological and computational vision. However, the existence of multiple motions in dynamic imagery due to occlusion or even transparency still raises challenging questions. In this paper, we propose an approach for the detection and characterization of occlusion and transparency. We propose a theoretical framework for both types of multiple motions which explicitly shows the difference between occlusion and transparency in the frequency domain. Then, we employ an EM-algorithm for the computation of one or two image velocities and a simple test for the detection of occlusion. Our approach differs from other EM-approaches which blindly assume the superposition of two models in the spatial domain without providing with a separate formal model for occlusion. We test and compare the characterization performance on synthetic and real data.

Real-time vehicle detection and tracking using stereo vision and multi-view AdaBoost

We propose a multi-layer, real-time vehicle detection and tracking system using stereo vision, multi-view AdaBoost detectors, and optical flow. By adopting a ground plane estimate extracted from stereo information, we generate a sparse set of hypotheses and apply trained AdaBoost classifiers in addition to fast disparity histogramming, for Hypothesis Verification (HV) purposes. Our tracking system employs one Kalman filter per detected vehicle and motion vectors from optical flow, as a means to increase its robustness. An acceptable detection rate with few false positives is obtained at 25 fps with generic hardware.

Petri Net-Based Cooperation In Multi-Agent Systems

We present a formal framework for robotic cooperation in which we use an extension to Petri nets, known as workflow nets, to establish a protocol among mobile agents based on the task coverage they maintain. Our choice is motivated by the fact that Petri nets handle concurrency and that goal reachability can be theoretically established. We describe the means by which cooperation is performed with Petri nets and analyze their structural and behavioral characteristics in order to show the correctness of our framework.

Modelling and Removing Radial and Tangential Distortions in Spherical Lenses

Spherical cameras are variable-resolution imaging systems and promising devices for autonomous navigation purposes, mainly because of their wide viewing angle which increases the capabilities of vision-based obstacle avoidance schemes. In addition, spherical lenses resemble the primate eye in their projective models and are biologically relevant. However, the calibration of spherical lenses for Computer Vision is a recent research topic and current procedures for pinhole camera calibration are inadequate when applied to spherical lenses. We present a novel method for spherical-lens camera calibration which models the lens radial and tangential distortions and determines the optical center and the angular deviations of the CCD sensor array within a unified numerical procedure. Contrary to other methods, there is no need for special equipment such as low-power laser beams or non-standard numerical procedures for finding the optical center. Numerical experiments, convergence and robustness analyses are presented.

Oriented structure of the occlusion distortion: is it reliable?

In the energy spectrum of an occlusion sequence, the distortion term has the same orientation as the velocity of the occluding signal. Other works claimed that this oriented structure can be used to distinguish the occluding velocity from the occluded one. We argue that the orientation structure of the distortion cannot always work as a reliable feature due to the rapidly decreasing energy contribution. This already weak orientation structure is further blurred by a superposition of distinct distortion components. We also indicate that the superposition principle of Shizawa and Mase (1991) for multiple motion estimation needs to be adjusted.

Workflow Nets for Multiagent Cooperation

We present a formal framework for robotic cooperation in which we use an extension to Petri nets, known as workflow nets, to establish a protocol among mobile agents based on the task coverage they maintain. Our choice is motivated by the fact that Petri nets handle concurrency and that goal reachability, or soundness, can be theoretically established. In particular, we define a mathematical cooperation operator which turns cooperation problems expressed as workflow nets into algebraic representations. While we do not address the problem of efficiency, we formally demonstrate that this framework guarantees soundness, or goal reachability, using workflow nets.

Portable and Scalable Vision-Based Vehicular Instrumentation for the Analysis of Driver Intentionality

Probably the most promising breakthroughs in vehicular safety will emerge from intelligent, Advanced Driving Assistance Systems (i-ADAS). Influential research institutions and large vehicle manufacturers work in lockstep to create advanced, on-board safety systems by means of integrating the functionality of existing systems and developing innovative sensing technologies. In this contribution, we describe a portable and scalable vehicular instrumentation designed for on-road experimentation and hypothesis verification in the context of designing i-ADAS prototypes.

An accurate discrete Fourier transform for image processing

The classical method of numerically computing the Fourier transform of digitized functions in one or in d-dimensions is the so-called discrete Fourier transform (DFT), efficiently implemented as Fast Fourier Transform (FFT) algorithms. In many cases the DFT is not an adequate approximation of the continuous Fourier transform. The method presented in this contribution provides accurate approximations of the continuous Fourier transform with similar time complexity. The assumption of signal periodicity is no longer posed and allows to compute numerical Fourier transforms in a broader domain of frequency than the usual half-period of the DFT. In image processing this behavior is highly welcomed since it allows to obtain the Fourier transform of an image without the usual interferences of the periodicity of the classical DFT. The mathematical method is developed and numerical examples are presented.