Comparison of Edge-Driven Algorithms for Model-Based Motion Estimation

Abstract

Abstract. 3D-model-based tracking offers one possibility to explicate the manner in which spatial coherence can be exploited for the analysis of image sequences. Two seemingly different approaches towards 3D-model-based tracking are compared using the same digitized video sequences of road traffic scenes. Both approaches rely on the evaluation of greyvalue discontinuities, one based on a hypothesized probability distribution function for step-discontinuities in the vicinity of model-segments, the other one based on extraction of Edge Elements (EEs) and their association to model-segments. The former approach could be considered to reflect a stronger spatial coherence assumption because the figureof-merit function to be optimized collects evidence from all greyvalue discontinuities within a tolerance region around visible model segments. The individual association of EEs to model-segments by the alternative approach is based on a distance function which combines differences in position and orientation, thereby taking into account the gradient direction as well as the location of a local gradient maximum in gradient direction. A detailed analysis of numerous vehicles leads to the preliminary conclusion that both approaches have different strengths and weaknesses. It turns out that the effects of how greyvalue discontinuities are taken into account are in general less important than the inclusion of Optical Flow (OF) estimates during the update-step of the current state vector for a body to be tracked. OF estimates are evaluated only within the area of the body to be tracked when projected into the image plane according to the current state estimate. Subtle effects related to simplifications and approximations during the implementation of an approach thus may influence the aggregated result of tracking numerous vehicles even in the case where spatial coherence appears to be rigorously exploited.

A Framework for Model-Based Tracking Experiments in Image Sequences

“…)-have been obtained using only standard parameter settings for the smaller vehicles (fastback, sedan, station-wagon).-The results presented above with the EEA and EMCo approaches are compatible to the ones reported inDahlkamp et al (2006), but they have been obtained here with a larger sample size and thus are statistically more relevant. -In combination with Optical Flow Association (OFA), both edge-based algorithms perform significantly better than without use of OF, resulting in a lead of OFEM over OFEE in the dt passat03 image sequence.…”

A Review of Computer Vision Techniques for the Analysis of Urban Traffic

“…This line of research is followed up in [57], [87], and [117]. A similar work using optical flow to find detection regions is presented in [109], with previous work in [32] and [33]. The 3-D wireframes of vehicles are used in a Hough transform to provide additional cues for vehicle detection.…”

Contour tracking based on marginalized likelihood ratios

“…It was initially applied to vehicle tracking with 3D models [44,42,45]. The vehicle tracker has been re-implemented at the University of Karlsruhe [7,8] and the Karlsruhe implementation is available online [61]. The MLR model has also been applied to real-time eye tracking with 2D models [17][18][19].…”