Transforming Traffic Management: Real-Time Vehicle Classification in Smart Transportation System

Vehicle classification can potentially alter intelligent transportation systems following recent advances in computer vision. Applications of Intelligent Transportation Systems improve traffic management by allowing us to make educated, safe, and smart decisions about transportation networks, resulting in increased safety and efficiency. Advances in computer vision and deep learning have improved vehicle classification systems. This study uses real-time data from Hyderabad City to propose a deep-learning model for recognizing random vehicle classes, including Light commercial vehicles (LCVs), Light motor vehicles (LMVs), Oversized vehicles (OSVs) and Trucks. Real-time data includes a selected number of images for training and testing. This work uses images of classes. Among them, the largest is LMV, and the least is OSVs. Every class has 30% testing and 70% training data. Preprocessing begins with resizing, scaling, and augmentation of the featured images. The CNN model was the base model, while Python libraries were used to train the network architecture. The validation accuracy was 88%, as per the test results.