The Study On The Law Of Influences Of High Texture Information Of Remote Sensing On The Measure Of The Forest Stock Volume Brought

Forest stock volume is one of the basic indicators to reflect the total size and level of forest resources in a country or region, is alsoan important basis to reflectthe richness of forest resources and to measure the merits of forest ecological environment. It's dynamic fluctuation not only directly affects the economic benefits of forestry, but also provides an important basis for the management of forests. So the investigation on the forest stock volume is quite necessary. The traditional stock volume estimation is estimated by collecting the position of the artificial surface acquisition of fixed size plots with a total area of data, and this method has heavy workload and low efficiency. With the development of 3S technology, now stock volume estimation is establishing stock volume estimation equation by combining survey data of fixed sample plots in the monitoring area with remote sensing data, to estimate the stock volume in forested areas. With increasingly sophisticated spatial resolution, the remote sensing image texture information has become increasingly diverse. High spatial resolution texture informationis very sensitive to the distinction of the forest vegetation information.When estimating the amount of accumulation, adding texture features has great potentialto enhance the accuracy of stock volume estimation.This paper is based on the image data from First High Point Road, using gray cooccurrence matrix to extract texture features of remote sensing image in testing areas, combining with remote sensing information and GIS information in testing areas to constructvolume estimation model,and, through experiments, exploringthe law of influences of texture features on the estimation of stand volume. We can draw conclusions as follow:(1) Using only the texture factor does cannot improve the accuracy of forest volume estimation;(2) Texture factor binding remote sensing and GIS factor can effectively enhance the accuracy of stock volume estimation;(3) Direction generating texture features can affect estimation accuracy of the model, but nobest direction was found tomake the generated texture feature to upgrade the accuracy of the model to the greatest extent;(4) With the increasing size of the window, the accuracy of model estimation shows a gradually decreasing trend, and the best texture extraction window appears in the 3 × 3 window.