In this work, the authors are attempting to improve the performance of Automatic Vehicle Recognition by employing Deep Convolutional Neural Network (DCNN). It is found that the algorithm performed best on real images and worse on synthetic images. It also emphasized the algorithm’s superiority over existing Optical Character Recognition (OCR) engines. The simulation results have been extracted with various performance measures based on F-Score. The proposed method has a detailed frame work for both training and test image. It has been implemented with existing standard database Tesseract and real time database sets. The existing database is not applicable to implement real moving vehicles. However, the proposed algorithm involves additional feature to extract the name plate details of the moving vehicles. This works based on deep convolutional network using recurrent process. It detects the number on the number plate after detecting the area of number plate. The proposed method gives better results based on confusion matrix parameters as compared the existing methods.
Moving Vehicle Number Plate Detection using Hybrid Deep Convolutional and Recurrent Neural Network Algorithm
04.08.2022
921947 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Moving vehicle detection using deep neural network
IEEE | 2017
|Emergency Vehicle Detection Using Deep Convolutional Neural Network
Springer Verlag | 2022
|Vehicle Number Plate Detection using Deep Learning
IEEE | 2024
|Expression of Concern for: Moving vehicle detection using deep neural network
IEEE | 2017
|