The Effect of Topographic Correction on Forest Tree Species Classification Accuracy

Topographic correction significantly enhances the accuracy of forest tree species classification using remote-sensing data, with the SCS+C model and empirical rotation model outperforming others. These models reduce band standard deviation, adjust reflectance distribution, and improve image quality, particularly in shaded areas. While total accuracy increased by 4% with full-coverage training areas and nearly 13% with shadowless training areas when using the SCS+C model, inconsistency rates decreased notably within an illumination condition interval of 0.4–0.6 but rose with higher values. The study, conducted at Mount Taishan forest farm using Landsat 8 OLI data and random forest classification, demonstrates that topographic correction is a valuable pretreatment method for classifying dominant tree species in low-light intensity areas.

Dong, C., Zhao, G., Meng, Y., Li, B., & Peng, B. (2020). The Effect of Topographic Correction on Forest Tree Species Classification Accuracy. Remote Sensing, 12(5), 787. https://doi.org/10.3390/rs12050787