AI-Enhanced Map-Matching Algorithm for Distinguishing Vehicular Movement on Highways and Service Roads with Deep Learning Techniques

The advancement of artificial intelligence (AI) and Deep learning (DL) techniques has created new opportunities for enhancing map-matching algorithms, particularly in distinguishing vehicular movement between highways and service roads. This study presents a novel map-matching approach that leverages Artificial and Deep Learning algorithms to improve the accuracy and reliability of identifying road types based on GPS data. By combining machine learning models with traditional map-matching techniques, the proposed method enhances the classification of vehicle trajectories, differentiating between high-speed highway travel and lower-speed service road navigation. The deep learning algorithm utilizes feature extraction, pattern recognition, and supervised learning to analyze GPS trajectories, integrating real-time data for dynamic and adaptive matching. Experimental results show that the Artificial Intelligence and Deep learning-based approach improves classification accuracy, achieving 84-90% for using RNN, 90–92% using CNN for GPS trajectory classification and 95–98% using YOLO for visual road identification. When combined, the hybrid model attains an overall accuracy of 96%, demonstrating its effectiveness in distinguishing between highway and service road travel.