Abstract: Using natural Pinus kesiya var. langbianensis forests in the southern subtropical region that serve as the study object, and we integrated UAV-borne Laser Scanning (ULS) and Backpack Laser Scanning (BLS) point clouds data to extract gap characteristics and calculate stand spatial-structure parameters, and then developed an optimization model driven by stand spatial structure indices. Subsequently, enrichment planting simulations and optimization assessments were conducted according to gap size classes. The results showed that individual tree attributes extracted from fused point clouds were highly consistent with field measurements (DBH: R² = 0.93, RMSE = 2.37 cm, Height: R² = 0.85, RMSE = 3.02 m). After enrichment planting, stand spatial-structure evaluation indices improved by 13.31%–125.01%, with an average increase of 65.31%. Moreover, the medium and large gaps showed average improvements of 75.05% and 75.17%, respectively. Mixing degree and openness generally increased, size differentiation decreased, angular scale approached random distribution, and competition indices showed an overall increasing trend. Overall, the integration of ULS and BLS point clouds enables fine-scale detection of canopy gaps, and enrichment-planting optimization using Genetic Algorithm (GA) can significantly improve stand spatial structure, with medium-sized gaps exhibiting more stable and higher management benefits.