Feature extraction analysis, techniques and issues in vehicle types recognition

Nor'aqilah, Misman and Suryanti, Awang (2018) Feature extraction analysis, techniques and issues in vehicle types recognition. In: Proceedings Book: National Conference for Postgraduate Research (NCON-PGR 2018), 28-29 August 2018 , Universiti Malaysia Pahang, Gambang, Pahang. pp. 28-35.. ISBN 978-967-22260-5-5

Feature Extraction Analysis, Techniques5.pdf

Download (688kB) | Preview


The Vehicle Type Recognition is one of the applications in the Intelligent Transportation System that has implemented in wide range areas such as intelligent parking systems and automatic toll collection system. The system is recognized and classified the vehicle based on vehicle types such as car, bus and truck classes. Most of the system’s accuracy depends on the features which represent the information from the data and the process of feature extraction whether to use single features extraction technique, a combination of single features techniques or based on deep learning methods. However, this paper focuses on feature extraction technique based on deep learning which is a Convolutional Neural Network. There are issues in the system that limit the capability which caused by overfitting, underfitting and intra-class issues. The intra-class issue occurs due to lack of features data and imbalanced dataset which is used for vehicle type classification. It happens when the recognition is applied to the vehicles with the almost similar appearance of the vehicle structure, for different vehicle type classes. To conclude, this paper discusses the related findings based on feature extraction techniques and issues in Vehicle Type Recognition; and used for a further research study to learn more about deep learning methods and data augmentation technique to improve the vehicle recognition and type classification especially in intra-classes.

Item Type: Conference or Workshop Item (Lecture)
Uncontrolled Keywords: Vehicle Type Recognition; Feature Extraction Technique; Convolutional Neural Network
Subjects: Q Science > QA Mathematics > QA76 Computer software
Faculty/Division: Faculty of Computer System And Software Engineering
Depositing User: Pn. Hazlinda Abd Rahman
Date Deposited: 10 Dec 2018 03:55
Last Modified: 24 Jul 2019 01:54
URI: http://umpir.ump.edu.my/id/eprint/23035
Download Statistic: View Download Statistics

Actions (login required)

View Item View Item