Description: We implement a system for vehicle detection and
tracking from traffic video using Gaussian mixture models and
Bayesian estimation. In particular, the system provides robust
foreground segmentation of moving vehicles through a K-means
clustering approximation as well as vehicle tracking correspon-
dence between frames by correlating Kalman and particle filter
prediction updates to current observations through the solution
of the assignment problem. In addition, we conduct performance
and accuracy benchmarks that show about a 90 percent reduc-
tion in runtime at the expense of reducing the robustness of
the mixture model classification and about a 30 percent and 45
percent reduction in accumulated error of the Kalman filter and
particle filter respectively as compared to a system without any
prediction.
To Search:
File list (Check if you may need any files):
TrafficDetection-master
TrafficDetection-master\Traffic_Detection_Estimation_Theory.pdf
TrafficDetection-master\addModel.m
TrafficDetection-master\centroidMatching.m
TrafficDetection-master\colorBox.m
TrafficDetection-master\connectedComponentCleanup.m
TrafficDetection-master\findNewLabel.m
TrafficDetection-master\foregroundEstimation.m
TrafficDetection-master\hungarian.m
TrafficDetection-master\initColorBoxes.m
TrafficDetection-master\initMatch.m
TrafficDetection-master\justCarsTrim.avi
TrafficDetection-master\matchingCriterion.m
TrafficDetection-master\matchingCriterionProb.m
TrafficDetection-master\particleMeasureUpdate.m
TrafficDetection-master\particleTimeUpdate.m
TrafficDetection-master\readMe.txt
TrafficDetection-master\removeModel.m
TrafficDetection-master\segmentBackground.m
TrafficDetection-master\vehicleTracking.m
TrafficDetection-master\vehicleTrackingParticle.m
Download