Pavement distress detection using terrestrial laser scanning point clouds – Accuracy evaluation and algorithm comparison

In this paper, we compared five crack detection algorithms using terrestrial laser scanner (TLS) point clouds. The methods are developed based on common cloud processing knowledge in along- and across-track profiles, surface fitting or local pointwise features, with without machine learning. area volume were calculated from the points detected by algorithms. completeness, correctness, F1 score of each algorithm computed against manually collected references. Ten 1-m-by-3.5-m plots containing 75 distresses six distress types (depression, disintegration, pothole, longitudinal, transverse, alligator cracks) selected to explain variability a 3-km-long-road. For at plot level, best achieved completeness up 0.844, correctness 0.853, an 0.849. algorithm’s overall (ten combined) 0.642, 0.735, 0.685 respectively. estimation, mean absolute percentage errors (MAPE) two 19.8% 20.3%. resulted 19.3% 14.5% MAPE. When grouped complexity, ‘easy’ category, reached estimation MAPE 8.9%, while for obtained was 0.7%.