A novel approach to measure and deconstruct pavement deformation and texture in the wheel track

Abstract

Conventional pavement evaluation methods struggle to resolve microscale texture degradation and early-stage deformations, leading to unreliable skid resistance assessment and premature maintenance interventions. This study proposed a hybrid spectral-spatial deconstruction method including three processes: planar inclination correction to eliminate geometric bias in raw three-dimensional laser-scanned data, frequency-domain decomposition to isolate deformation (low-frequency) and texture (high-frequency) components, and bidirectional second-order polynomial fitting with directional regularization data for deformation surface reconstruction. The framework established four deformation indices: absolute/relative maximum depth deviation, deformation-affected area, and rut width limit. Seven in-service roads with gradient wear and rutting were detected to validate the tilt calibration, spectral filtering, and bi-directional fitting of the individual processing steps, demonstrating the efficacy of the proposed method. It was proved that the rut depth metrics proposed can effectively characterize rut geometry across asymmetric deformation scenarios. Mean texture depth stability was significantly enhanced through the proposed workflow under mild deformation. This methodology redefines deformation-texture analysis paradigms, offering a critical tool for pavement condition monitoring. While preliminary results validate the method’s effectiveness, extended field trials are warranted for broader generalization.