| Canopy density is one of the important factors stand survey.lt is an important indicator of the merits of forest resources.Use of remote sensing data, combined with ground survey data to achieve estimates canopy density is one of the research of remote sensing data for forest resources survey.In this paper,the Genhe forestry Chaocha state forest farm of Inner Mongolia Greater Khingan Mountains as the study area, with QuickBird data and Forest Resource Inventory data for research data.ResearchData integration and shadow elimination, the associated canopy density estimation method based on remote sensing image crown extraction and canopy density regression model was constructed to estimate of the QuickBird image data.The main conclusions of the study are as follows:(1) Using HSV fusion, Gram-Schmidt fusion, principal component transformation, Pan sharpening multi-spectral image fusion and integration of four panchromatic image fusion methods, evaluate the effectiveness of the best fusion method is Pan sharpening.(2) For image fusion using histogram equalization, Gamma correction, homomorphic filtering, Retinex enhanced four ways to eliminate the shadow of the trees in the study area, the present study the most appropriate method to eliminate the shadow of homomorphic enhancement.(3) Canopy density estimation based on the extracted image crown canopy density estimating method, compared with the Forest Resource Inventory Data Found canopy density, estimation accuracy of 86.15%, reached the C-class.(4) Optimal canopy density estimation model is:Y=0.456279-0.007647X1-0.002585X2+0.011410X3+0.61313X4 Estimation accuracy is 87.14%.(5) Canopy density estimation based on the average estimate crown extraction accuracy is 86.15%.Optimal Modeling canopy density estimation model average accuracy is 87.14%.Modeling canopy density estimation average accuracy is higher than canopy density estimation average accuracy based on the crown extraction. |
