Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Deep Learning-Based Models with YOLOv7 and Convolutional Neural Networks for Vehicle Detection and Recognition

Rektorat UDINUS 2025 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2025
Goals
Record
10b71f40-9ff3-4ae3-aca8-1d40ed1e7b43
Published
2026-10-30

Universitas Dian Nuswantoro conducted research on the application of artificial intelligence technology in transportation and smart city contexts. The activity involved Nugroho, Wahyu Adi; Supriyanto, Catur; and Safar, Noor Zuraidin Mohd. The research addressed limitations in traffic monitoring approaches that rely on license plate recognition, particularly when plates are obscured. It explored recognition of broader vehicle attributes to support vehicle identification and comprehensive traffic statistics.

The proposed solution integrated the You Only Look Once object detection algorithm with pretrained convolutional neural network models for vehicle attribute recognition. The study used the Roboflow Vehicle, Stanford Cars, VehicleID, and VCoR datasets for model evaluation. The evaluation covered vehicle detection, vehicle make and model recognition, and vehicle color recognition. YOLOv7 achieved a mean average precision score of 86.1% for vehicle detection and an average precision score of 91.5% for the car class. For vehicle make and model recognition, the lightweight EfficientNetV2S model achieved accuracy scores of 89.8% on the Stanford Cars dataset and 99.2% on the VehicleID dataset. For vehicle color recognition, DenseNet201 achieved an accuracy of 87% on the VCoR dataset.

The reported findings indicate the effectiveness of integrating YOLOv7 and convolutional neural network models for vehicle detection and recognition. The research presents a broader vehicle-attribute approach to limitations associated with traditional license plate recognition. It also identifies further optimization for real-time applications and diverse traffic scenarios as a possible direction for future studies. The activity is institutionally framed within research addressing vehicle recognition and traffic monitoring in transportation and smart city contexts.

Reference https://www.scopus.com/pages/publications/105020730885?origin=resultslist