Douglas-fir (Pseudotsuga menziesii) is a highly valued lumber for heavy duty construction which is exceptionally hard, stiff and durable. However, it has a long growth circle and very sensitive to the environmental changes, particularly to the ground treatments. In this study, we try to use remote sensing technique to assess the Douglas fir productivity under different ground treatments. We developed a two stages approach. In the species classification stage, we applied an innovative crown pattern recognition approach on very high spatial resolution imagery to identify the individual Douglas-fir. We then used regression analysis to explore the correlation between ground treatment and tree productivity.