A systematic review of the use of computer vision and photogrammetry tools in learning-based dimensional road pavement defect detection for smart transportation

Abstract

This study presents a comprehensive literature review focused on the utilization of computer vision, photogrammetry tools, and machine learning algorithms to enhance road pavement condition assessment and contribute to the advancement of smart transportation trends. It conducts a systematic review of literature focused on road pavement condition assessment, specifically leveraging computer vision, photogrammetry tools, and machine learning algorithms to promote smart transportation. The research methodically extracts and analyzes pertinent literature from reputable sources, resulting in the identification of 56 articles out of an initial pool of 136. The review encompasses diverse aspects, such as the use of computer vision and photogrammetry tools, model functionalities, infrastructure types, data acquisition methods, and software tools. It also scrutinizes the challenges posed by these techniques, identifies research gaps, and considers their potential implications for the integration of autonomous vehicles and smart transportation. Key findings stress the necessity for standardizing evaluation parameters, achieving real-time applicability, and aligning with the objectives of smart transportation. Ultimately, this review underscores the urgency of further research in these domains to advance the development of smart transportation solutions.