Abstract: This work proposes a comprehensive approach utilizing the YOLOv8 deep learning algorithm to enhance vehicle detection and classification in intelligent transportation systems (ITS). The methodology involves meticulous dataset preparation, including diverse traffic videos and pre-processing techniques to ensure dataset quality. Leveraging the YOLOv8 algorithm implemented through the Ultralytics framework, the model is fine-tuned using transfer learning on custom datasets. Results demonstrate the effectiveness of the YOLOv8 model in accurately detecting and classifying vehicles, with further enhancements achieved through model optimization techniques like hyperparameter tuning and post-processing methods. The findings contribute to advancing computer vision and deep learning applications in transportation, paving the way for improved traffic management systems and autonomous vehicle technology.

Keywords: Vehicle Detection, Vehicle Identification, YOLOv8, Deep Learning, Intelligent Transportation Systems.

| DOI: 10.17148/IARJSET.2024.11456