NORMA eResearch @NCI Library

An Efficient Deep Neural Network For Traffic Sign Classification in Autonomous Vehicles

Sebastian Varghese, Aiswarian (2022) An Efficient Deep Neural Network For Traffic Sign Classification in Autonomous Vehicles. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
PDF (Master of Science)
Download (1MB) | Preview
[thumbnail of Configuration manual]
PDF (Configuration manual)
Download (1MB) | Preview


Intelligent transportation system is an emerging leap in the automobile sector, Classifying and recognizing the road and traffic signs are the supreme aspect for the companies for developing their own prototype in this industry. Convolutioal neural networks in deep learning is an fortunate method to attain accurate results in these image processing tasks.

Although Numerous approach by different researchers have achieved promising results in classifying and predicting traffic signs using traditional hand crafted methods, The proposed work focuses on developing a convolutional neural network classifier where this deep learning technique classifies all the traffic signs images with improved accuracy and precision. the performance of the model is further improved by effective fine tuning in the hyper parameters. The data points in the considered dataset contained the traffic sign images and markings in different environmental conditions and distinct lighting conditions. These signs are first preprocessed to identify its extract clinical information from it and then preprocessed images are transformed to identify the area of interest thus signs are accurately recognized using deep learning. We use an end to end open source platform tensor flow to implement CNN. The hyperparameter tuning of the model is performed using Grid search CV technique and optimized value for the parameters are chosen for improved efficiency of the model. The performance of the model is evaluated using the performance metrics. The presented model gaid the overall accuracy of 98.44% using the deep neural network model.

Item Type: Thesis (Masters)
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
T Technology > TE Highway engineering. Roads and pavements
T Technology > TL Motor vehicles. Aeronautics. Astronautics
Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Tamara Malone
Date Deposited: 10 Mar 2023 17:43
Last Modified: 10 Mar 2023 17:43

Actions (login required)

View Item View Item