AI SearchAutomatic vehicle identification and matching in multiple perspectives
Somphop Limsoonthrakul
Automatic vehicle identification and matching in multiple perspectives - Pathum Thani, Thailand : Asian Institute of Technology, 2021 - 46 leaves : ill. - Dissertation ; no. ISE-21-03 . - Asian Institute of Technology. Dissertation ; no. ISE-21-03 .
A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Engineering in Mechatronics and Embedded Systems, School of Engineering and Technology
Thesis (Ph. D.) - Asian Institute of Technology, 2021
This dissertation propose an implementation of automated vehicle tracking system based on vision sensors and video analytics. The proposed system can process video streams from multiple traffic cameras to identify vehicles that appears in the scenes and predicts the route of a particular vehicle by matching the visual properties (type, color, and make) and/or license number found in adjacent cameras. Classical image processing techniques and Convolutional Neural Network architectures (GoogLeNet and YOLO) are adopted for vehicle detection-classification, and license plate recognition. The study also propose an architectural design of distributed traffic camera system which can reduce the cost of installation in wide coverage area.
Automobiles--Identification
Automated vehicles
Image processing--Digital techiniques
Automatic vehicle identification and matching in multiple perspectives - Pathum Thani, Thailand : Asian Institute of Technology, 2021 - 46 leaves : ill. - Dissertation ; no. ISE-21-03 . - Asian Institute of Technology. Dissertation ; no. ISE-21-03 .
A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Engineering in Mechatronics and Embedded Systems, School of Engineering and Technology
Thesis (Ph. D.) - Asian Institute of Technology, 2021
This dissertation propose an implementation of automated vehicle tracking system based on vision sensors and video analytics. The proposed system can process video streams from multiple traffic cameras to identify vehicles that appears in the scenes and predicts the route of a particular vehicle by matching the visual properties (type, color, and make) and/or license number found in adjacent cameras. Classical image processing techniques and Convolutional Neural Network architectures (GoogLeNet and YOLO) are adopted for vehicle detection-classification, and license plate recognition. The study also propose an architectural design of distributed traffic camera system which can reduce the cost of installation in wide coverage area.
Automobiles--Identification
Automated vehicles
Image processing--Digital techiniques