ech T Press Science Computers, Materials & Continua DOI:10.32604/cmc.2021.015504 Article An Optimized Approach to Vehicle-Type Classifcation Using a Convolutional Neural Network Shabana Habib 1 and Noreen Fayyaz Khan 2, * 1 Department of Information Technology, College of Computer, Qassim University, Buraidah, 51452, Saudi Arabia 2 Department of Computer Science, Islamia College University, Peshawar, Pakistan * Corresponding Author: Noreen Fayyaz Khan. Email: noreen.fayyaz@fu.edu.pk Received: 25 November 2020; Accepted: 02 March 2021 Abstract: Vehicle type classifcation is considered a central part of an intel- ligent traffc system. In recent years, deep learning had a vital role in object detection in many computer vision tasks. To learn high-level deep features and semantics, deep learning offers powerful tools to address problems in traditional architectures of handcrafted feature-extraction techniques. Unlike other algorithms using handcrated visual features, convolutional neural net- work is able to automatically learn good features of vehicle type classifcation. This study develops an optimized automatic surveillance and auditing system to detect and classify vehicles of different categories. Transfer learning is used to quickly learn the features by recording a small number of training images from vehicle frontal view images. The proposed system employs extensive data- augmentation techniques for effective training while avoiding the problem of data shortage. In order to capture rich and discriminative information of vehicles, the convolutional neural network is fne-tuned for the classifcation of vehicle types using the augmented data. The network extracts the feature maps from the entire dataset and generates a label for each object (vehicle) in an image, which can help in vehicle-type detection and classifcation. Experi- mental results on a public dataset and our own dataset demonstrated that the proposed method is quite effective in detection and classifcation of different types of vehicles. The experimental results show that the proposed model achieves 96.04% accuracy on vehicle type classifcation. Keywords: Vehicle classifcation; convolutional neural network; deep learning; surveillance 1 Introduction Surveillance systems have achieved good results in terms of security. Image analysis, such as detecting a moving vehicle in an image, is a challenging task that can be solved by analyzing the foreground [1]. Dramatic improvements have been observed in the areas of speech recognition and document recognition genomics for automation technologies [2]. Major issues in surveillance systems include brightness, lighting, occlusion of shadows, and fragmentation, and all have a negative impact on objects to be detected [3,4]. Much research has been done on vehicle-type This work is licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.