Abstract: In this study, the natural forest of Pinus kesiya var. langbianensis was taken as the research subject, and tree height (H), diameter at breast height (DBH), species, and location data of individual trees were extracted using the three-dimensional spatial information obtained from Backpack Laser Scanning (BLS). The accuracy of the BLS data was validated using sample plot data. This information was then used to calculate stand spatial structure parameters. Furthermore, the entropy weight method was employed to assign weights to these parameters, and a thinning model for optimizing the stand spatial structure was developed by integrating the principle of multiplication and division. This model was then used to determine the optimal thinning intensity. The results showed that the accuracy of DBH extraction using BLS has an R² of 0.90 and an RMSE of 2.19 cm, while the accuracy of H extraction has an R² of 0.77 and an RMSE of 2.09 m. Among the 5 stand spatial structure parameters, the mixing degree had the highest weight (48.12%), while the competition index had the lowest (3.08%). The optimal thinning intensity was determined to be 15%, resulting in an 81.89% increase in the stand spatial structure evaluation index and improvements in all stand structure parameters after thinning. This study demonstrates that the thinning model constructed based on BLS data collection, integrating the entropy weighting method and the principles of multiplication and division, can effectively evaluate and optimize the stand spatial structure.