Laser scanning technology is widely used to characterize the surface texture of asphalt pavements. However, raw point-cloud data are susceptible to noise, isolated spikes, and nonuniform sampling, which may reduce the accuracy of texture characterization. To address these issues, a rapid workflow for point-cloud processing, correction, and visualization was developed. The workflow consisted of outlier screening based on the 3σ criterion, regular gridding, bilinear interpolation, and elevation datum correction through mean centering. Subsequently, the sensitivities of six texture indicators to abrasion duration were evaluated using linear regression, relative change analysis, and Spearman’s rank correlation coefficient. On this basis, a texture characterization system comprising mean profile depth (MPD) and arithmetic mean height (Ra) was established, and its reliability and validity were verified through regression analysis against the British Pendulum Number (BPN). The proposed indicator system was then applied to evaluate the evolution of pavement surface texture under different abrasion durations and ultraviolet aging conditions. The results indicated that the texture indicators exhibited a stage-dependent evolution characterized by an initial decrease, a temporary in-crease, and a subsequent decrease during abrasion. This behavior was associated with the varying contributions of the asphalt binder film and exposed aggregates to the surface texture at different abrasion stages. After ultraviolet aging, the degradation rates of MPD and Ra increased by 19.3% and 12.5%, respectively, indicating that ultraviolet aging accelerated the deterioration of pavement surface texture. The proposed method provides methodological support and a reliable data basis for the accurate characterization of asphalt pavement texture and the evaluation of its degradation behavior.