A pavement distress survey algorithm with novel models and line points detection

With the fast development of the public transportation and logistics, the automatic distress detection is necessary to save the time, cost and labor. This paper introduces a pavement distress automatic detection algorithm. Its main idea is based on the fact that the distress in the pavement can be modeled as the curve segments or curve segment networks. Three steps are necessary. First of all, the noise is suppressed based on the proposed pavement and noise models, then the potential line points are detected, and at last, the line points are linked to form the curve segments or curve segment networks. The experiment tests show that the proposed algorithm can correctly detect and precisely locate the distress at the same time.