Efficient image gradient based vehicle localization

Abstract

This paper reports novel algorithms for the efficient localization and recognition of traffic in traffic scenes. The algorithms eliminate the need for explicit symbolic feature extraction and matching. The pose and class of an object is determined by a form of voting and one-dimensional (1-D) correlations based directly on image gradient data, which can be computed "on the fly." The algorithms are therefore very well suited to real-time implementation. The algorithms make use of two a priori sources of knowledge about the scene and the objects expected: (1) the ground-plane constraint and (2) the fact that the overall shape of road vehicles is strongly rectilinear. Additional efficiency is derived from making the weak perspective assumption. These assumptions are valid in the road traffic application domain. The algorithms are demonstrated and tested using routine outdoor traffic images. Success with a variety of vehicles in several traffic scenes demonstrates the efficiency and robustness of context-based image understanding in road traffic scene analysis. The limitations of the algorithms are also addressed in the paper.

Explicit contour model for vehicle tracking with automatic hypothesis validation

“…Their method, however, was limited to a single class of vehicle only and assumed that all vehicles move in fixed directions. Tan et al [6] proposed an efficient algorithm which makes use of 3D models to estimate the shapes of vehicles at different poses and locate them in an image. To recover the 3D pose from a single image, Tan's work requires the ground plane constraint, which, however, may not hold at curved or curly roads.…”

A Survey on Visual Surveillance of Object Motion and Behaviors

“…Tan et al [87], [119] propose the ground-plane constraint (GPC), under which vehicles are restricted to move on the ground plane. Thus the degrees of freedom of vehicle pose are reduced to three from six.…”

Robust Vehicle Registration Method Based on 3D Model for Traffic Surveillance Application

“…The intensity values of image pixels around model projection are utilized in [13]. Kollnig and Nagel [14] generated a synthetic gradient by convolving model projection with a Gaussian noise and then compared this gradient with image gradient.…”

“…In recent years, 3D-2D vehicle registration has attracted more and more attentions, which provides a new way for vehicle recognition [8][9][10][11], localization [12][13][14] and tracking [15][16][17][18]. With the rapid development of 3D modeling technology, 3D vehicle model can be easily obtained.…”