Our study explores the spatiotemporal mechanisms by which three-dimensional (3D) canopy characteristics affect human thermal perception across seasons. Through environmental monitoring and spatial analysis, we assessed key canopy parameters (mean foliage height [MFH], foliage height diversity [FHD], crown width [CW], and leaf area index [LAI]) along pedestrian pathways in an urban park in Xi'an, China. A portable backpack LiDAR was employed to capture high-resolution 3D canopy data, analyzed via concentric buffers (5 m, 10 m, 15 m). Meteorological data was recorded with mobile instruments, and thermal perceptions were surveyed via questionnaires. Results showed significant spatial heterogeneity in canopy structure, with 5 m buffers exhibiting greater variation. Strong seasonal correlations existed between canopy characteristics and microclimate parameters, particularly for air temperature (T-a) and relative humidity (RH). The downward shortwave radiation (Q(sw-d)) dominated thermal sensation vote (TSV) in autumn, downward longwave radiation (Q(lw-d)) in winter. While T-a, RH, and Q(sw-d) correlated highly with TSV in spring and summer (p < 0.05). CW had a more direct impact on TSV in spring and autumn.MFH was the key canopy predictor of TSV in winter and summer. LAI and FHD primarily influenced thermal sensation by modulating Q(sw-d), Q(lw-d), and RH. In summer, canopies with taller MFH and moderate LAI were most effective in blocking direct solar radiation and enhancing ventilation. In spring and autumn, wider CW combined with moderate FHD provided optimal shading and enhanced psychological perception of comfort. In winter, moderate MFH and relatively low FHD reduced longwave radiation loss and shielded against wind, enhancing perceived warmth. Our results offer science-based guidance to prioritize tree species with seasonally appropriate CW, MFH, and moderate LAI/FHD combinations to enhance thermal comfort and mitigate urban heat islands.