The Application Of SPOT5 Remote Sensingdata In Forestry Surveying

"3S" technology-based methods for forest resources surveying has become the main direction of development of update forest resources surveying technology. Probe for appropriate methods of forest resources surveying, especially the forest classification by remote sensing data and the sub-compartment volume estimation, has become the pressing issues.This study used SPOT5 data with 2.5m resolution for forest resources surveying, thoroughly discussed the methods of SPOT5 remote sensing data processing, remote sensing classification of forests and sub-compartment volume remote sensing estimation. We attempt to provide the scientific basis and supports of technical methods for forest resources surveying by SPOT5 data. Major research results are such as follows:(1) Put forward the SPOT5 data processing methods for forest resources surveying in south china. In practical applications, are advised to use the following technical processes : Collect a sufficient number of ground control points by high-precision WADGPS; Complete the panchromatic and multispectral data correction and spatial matching with geometric polynomial correction associate;Complete fusion of data using IHS transformations; Combine with the 1:10000 DEM data and SPOT5 physical model to do the image Orthorectification;Do the spatial enhancement by nonlinear stretching and convolution; XS2, (XS1*3+XS3) /4 and XS1 respectively delegate to band red ,green and Blue band to achieve the true-colour combination images.(2) On the basis of SPOT5 spectral and texture remote sensing data, information of historical forest surveying data and experts classification system, put forward a new classification model. The classification model has a good results, the overall classification accuracy reached 92.97 %, Kappa factor of 0.9172, and all types classification accuracy are achieved 87% or more. Historical surveying data in this classification can effectively improve their classification accuracics, its overall contribution rate is 11.55%. Based on the classification, completed the extraction of sub-compartment boundary and area estimation effectively.(3) Discussed the method of forest volume per hectare estimation in depth. Based on the analysing of the data, selected relevant factors, made a series of tests and amendments with models, then created forest volume estimation optimal multivariate linear regression model. Combine with Estimation values of the sub-compartment area, completed the forest volume estimation.