Thesis Details
Reidentifikace automobilů ve videu
This thesis deals with the vehicle re-identification in video problem. Vehicle re-identification is based on matching image parts obtained from different cameras. This work is focues on the re-identification itself assuming that the vehicle detection problem is already solved including extraction of a full-fledged 3D bounding box. The re-identification problem is solved by using color histograms, histograms of oriented gradients by a linear regressor. The features are used in separate models in order to get the best results in the shortest CPU computation time. The proposed method works with a high accuracy (60% true positives retrieved with 10% false positive rate on a challenging subset of the test data) in 85 milliseconds of the CPU (Core i7) computation time per one vehicle re-identification assuming the Full HD resolution video input. The applications of this work include finding important parameters like travel time, traffic flow, or traffic information in a distributed traffic surveillance and monitoring system.
vehicle re-identification, anonymous re-identification, computer vision, regression, feature extraction, image matching, machine learning
Janoušek Vladimír, doc. Ing., Ph.D. (DITS FIT BUT), člen
Rychlý Marek, RNDr., Ph.D. (DIFS FIT BUT), člen
Strnadel Josef, Ing., Ph.D. (DCSY FIT BUT), člen
Szőke Igor, Ing., Ph.D. (DCGM FIT BUT), člen
@bachelorsthesis{FITBT17333, author = "Dominik Zapletal", type = "Bachelor's thesis", title = "Reidentifikace automobil\r{u} ve videu", school = "Brno University of Technology, Faculty of Information Technology", year = 2015, location = "Brno, CZ", language = "czech", url = "https://www.fit.vut.cz/study/thesis/17333/" }