Národní úložiště šedé literatury Nalezeno 2 záznamů.  Hledání trvalo 0.01 vteřin. 
Detection and Recognition of License Plates
Tykva, Jiří ; Zemčík, Pavel (oponent) ; Juránek, Roman (vedoucí práce)
The goal of this Bachelor's thesis is to design, implement, and test a system that can detect and recognize license plates in real-time by using neural networks. The collected data will be saved into the database. The system's architecture is divided into three main parts. The first part handles the license plate detection in the image by making use of the TensorFlow Object Detection API. The detector reaches the accuracy of 98.15 % AP with a speed of roughly 14 fps. The second part deals with license plate tracking by using the algorithm SORT. The third part holistically recognizes the text of the license plate and can reach up to 0.6% character error rate and 2% word error rate. The system may be used by law enforcement for purposes such as for keeping track of stolen vehicles or for the automatic road tolling.
Detection and Recognition of License Plates
Tykva, Jiří ; Zemčík, Pavel (oponent) ; Juránek, Roman (vedoucí práce)
The goal of this Bachelor's thesis is to design, implement, and test a system that can detect and recognize license plates in real-time by using neural networks. The collected data will be saved into the database. The system's architecture is divided into three main parts. The first part handles the license plate detection in the image by making use of the TensorFlow Object Detection API. The detector reaches the accuracy of 98.15 % AP with a speed of roughly 14 fps. The second part deals with license plate tracking by using the algorithm SORT. The third part holistically recognizes the text of the license plate and can reach up to 0.6% character error rate and 2% word error rate. The system may be used by law enforcement for purposes such as for keeping track of stolen vehicles or for the automatic road tolling.

Chcete být upozorněni, pokud se objeví nové záznamy odpovídající tomuto dotazu?
Přihlásit se k odběru RSS.