Measurement of 3D pavement morphology using improved binocular reconstruction under structured light constraint mechanism

Traditional binocular vision is not ideal for measuring the three-dimensional (3D) morphology of asphalt pavements owing to the dark surface colour, weak regional texture, and inconspicuous feature points. To achieve accurate measurement of 3D pavement morphology, this study introduces sine-encoded stripe-structured light and explores a matching strategy that combines spatial-domain image information with frequency-domain phase information. A phase-shift iterative matching algorithm under a structured-light constraint mechanism is proposed. In addition, a binocular reconstruction measurement system is developed to evaluate the measurement accuracy and the measurement effect. The results show that the constraint effect of structured light enhances effective pixel matching and improves the measurement accuracy. The maximum absolute deviation measured by the proposed algorithm is only 0.4 mm, which can meet the precision requirements of 3D pavement morphology measurements. This study offers crucial insights for achieving a high-precision digital road model and creating a comfortable and safe traffic environment.