Methods for advanced traffic analysis in cloud
Herout Adam, prof. Ing., Ph.D. (DCGM FIT BUT)
vehicle re-identification; convolutional neural network; feature extraction; cloud processing
Traffic analysis software in the cloud, created within the MV SECTECH project - Safe and Secure Traffic Systems of New Generation
The web application for processing traffic data in the cloud offers an interface (REST API) for the task of re-identification of vehicles based on visual characteristics (appearance) of the vehicle in the image, together with the actual implementation of the developed methods.
This framework, based on Flask, RabbitMQ, Redis, Celery techniques, offers distributed traffic data processing capabilities, including computation of convolutional neural network models for extracting identification feautres using ONNX runtime.
The developed web application offers the possibility of extending to other traffic analysis tasks for further development of the project.
For research purposes as open-source, for commercial purposes under license.
For information about license conditions please contact: Research Centre of Information Technology, Faculty of Information Technology, Brno University of Technology, Božetěchova 2, 612 66 Brno, +420 541 141 472.