This paper presents an e cient shape-based object detection method based on Distance Transforms and describes its use for real-time vision on-board vehicles. The method uses a template hierarchy to capture the variety of object shapes; e cient hierarchies can be generated o ine for given shape distributions using stochastic optimization techniques i.e. simulated annealing. Online, matching involves a simultaneous coarse-to-ne approach o ver the shape hierarchy and over the transformation parameters. Very large speedup factors are typically obtained when comparing this approach with the equivalent brute-force formulation; we h a ve measured gains of several orders of magnitudes.We present experimental results on the real-time detection of tra c signs and pedestrians from a moving vehicle. Because of the highly time sensitive nature of these vision tasks, we also discuss some hardwarespeci c implementations of the proposed method as far as SIMD parallelism is concerned.

The surface profile histories of gentle spilling breakers generated mechanically with a dispersive focusing technique are studied experimentally. Froude-scaled generation conditions are used to produce waves with three average frequencies: f0=1.42, 1.26, and 1.15 Hz. At each frequency, the strength of the breaker is varied by varying the overall amplitude of the wavemaker motion. It is found that in all cases the beginning of the breaking process is marked by the formation of a bulge in the profile at the crest on the forward face of the wave. The leading edge of this bulge is called the toe. As the breaking process continues, the bulge becomes more pronounced while the toe remains in nearly a fixed position relative to the crest. Capillary waves form ahead of the toe. At a time of about 0.1/f0 after the bulge first becomes visible, the toe begins to move down the face of the wave and very quickly accelerates to a constant velocity which scales with the wave crest speed. During this phase of the breaker evolution, the surface profile between the toe and the crest develops ripples which eventually are left behind the wave crest. It is found that the height of the toe above the mean water level scales with the nominal wavelength λ0=g/(2πf20) of the breaker, while the size and shape of the bulge and the length of the capillary waves ahead of the toe are independent of f0.

Abstract. The problem of tracking pedestrians from a moving car is a challenging one. The Condensation tracking algorithm is appealing for its generality and potential for real-time implementation. However, the conventional Condensation tracker is known to have di culty with high-dimensional state spaces and unknown motion models. This paper presents an improved algorithm that addresses these problems by using a simplified motion model, and employing quasi-Monte Carlo techniques to e ciently sample the resulting tracking problem in the high-dimensional state space. For N sample points, these techniques achieve sampling errors of O(N 31 ), as opposed to O(N 31/2 ) for conventional Monte Carlo techniques. We illustrate the algorithm by tracking objects in both synthetic and real sequences, and show that it achieves reliable tracking and significant speed-ups over conventional Monte Carlo techniques.

Photographs from high-speed movies of the profiles of a mechanically generated, gentle spilling breaking water wave are presented. It is found that as the wave steepens a bulge forms on the forward face of the wave near the crest and capillary waves form on the water surface ahead of the ‘toe’ of the bulge (see Fig. 1). The toe of the bulge then moves rapidly down the forward face of the wave and a train of large-amplitude waves with short wavelength grows rapidly on the surface of the bulge. These waves quickly break down into a random pattern indicating that the flow has become turbulent.

No abstract