PROJECT TITLE :

Counting and Classification of Highway Vehicles by Regression Analysis

ABSTRACT:

During this paper, we tend to describe a novel algorithm that counts and classifies highway vehicles based on regression analysis. This algorithm needs no explicit segmentation or tracking of individual vehicles, which is usually an important half of many existing algorithms. So, this algorithm is particularly helpful when there are severe occlusions or vehicle resolution is low, in which extracted options are highly unreliable. There are mainly two contributions in our proposed algorithm. 1st, a warping method is developed to detect the foreground segments that contain unclassified vehicles. The common used modeling and tracking (e.g., Kalman filtering) of individual vehicles aren't required. So as to scale back vehicle distortion caused by the foreshortening result, a nonuniform mesh grid and a projective transformation are estimated and applied throughout the warping method. Second, we have a tendency to extract a group of low-level features for every foreground segment and develop a cascaded regression approach to count and classify vehicles directly, which has not been used in the realm of intelligent transportation systems. 3 completely different regressors are designed and evaluated. Experiments show that our regression-primarily based algorithm is correct and sturdy for poor quality videos, from which several existing algorithms might fail to extract reliable options.