Smart Vehicle Classification Framework for Traffic Analysis and Sustainable Mobility

Abstract

The rapid growth of vehicle imagery on digital platforms has increased the
demand for accurate and efficient vehicle classification systems, particularly for applications
such as traffic surveillance, autonomous driving, and intelligent transportation management.
In this study, we propose a Convolutional Neural Network (CNN)based multi-class
classification model capable of recognizing seven vehicle categories: Auto Rickshaws, Bikes,
Cars, Motorcycles, Planes, Ships, and Trains. The dataset consists of 5,007 JPG images, of
which 3,504 images are used for training and 1,503 for testing. The CNN architecture is
carefully designed to extract discriminative and hierarchical features unique to each vehicle
type. Extensive experimentation and hyperparameter tuning are conducted to optimize the
model’s performance. The results demonstrate the effectiveness of CNNs in vehicle
recognition tasks. Accurate vehicle classification contributes to improved traffic flow,
reduced congestion, and enhanced transportation efficiency, thereby supporting sustainable
urban development. Furthermore, the proposed system aligns with goals related to
responsible consumption and production, industrial innovation, infrastructure development,
and climate action by enabling smarter and more environmentally efficient transportation
solutions