森林蓄积量作为表征森林数量的重要指标之一，反映了一个国家或地区森林资源总规模和水平的基本指标之一，是制定计划采伐、森林经营管理的依据。随着GIS与遥感技术的快速发展和普及应用，采用高分辨率遥感影像结合少量地面调查样地进行森林蓄积量估测，已成为林业研究的热点。目前国内基于GIS和遥感影像进行蓄积量估测的研究主要是使用传统的多元线性回归方法，导致估测精度下降。本文将研究常用于蓄积量估测五种算法，建立并优选分类模型以提高精度和模型适用性。 蓄积量分类估测的前提是遥感图像可以有效的分为对应的类型，首先需对图像进行预处理，尽可能恢复目标的反射光谱特性和正确的几何位置；然后需对林地信息进行人工解译，建立解译标准；最后对遥感图像做监督分类和重编码。 为验证分类估测和Box-Cox变换的效果，本文使用北京市密云县第七次全国森林资源一类清查样地数据和经过几何校正的同期TM遥感影像及30米分辨率DEM数据作为基础数据，对北京密云县森林蓄积量进行分类估算，并采用Box-Cox对蓄积量进行修正变换，分析得出：蓄积量分类估测可以有效的提高估测精度和模型的预报能力，可以把分类建模有效的用于蓄积量估测中；在相同的建模因子、相同的建模样地的条件下，蓄积量Box-Cox修正后模型虽然精度略有降低，但可以很大提高模型的预报能力和适用性， 在做蓄积量估测时应该做回归诊断，在异常样地个数较多而影响Gauss-Markov假设时，可以引入Box-Cox变换。

Forest stock volume as an important indicator of the mount of forest that is one of the basic indicators of reflecting the total size of the forest resources and levels of a country or region ,is basis for planning deforestation and forest management. With the rapid development and application of GIS and remote sensing technology, it has become a hot forestry research topic to use high-resolution remote sensing images combined with limited ground survey plots in forest stock volume estimation. At present traditional multiple linear regression method is adopted by the domestic stock volume estimation based on GIS and remote sensing research, and results in decrease of estimation accuracy. This article will examine five common algorithms in stock volume estimation, create and prefer classification model to improve the accuracy and model applicability. The premise of stock volume estimation is that the classification of remote sensing images can be effectively divided into corresponding type. First, we need preprocess the images as much as possible to restore the target reflectance characteristics and the correct geometric position. Then we need interpret information manually on woodland and establish interpret standards. Finally, to make supervised classification of remote sensing images and to re-encode. In order to verify the effect of classification estimation and Box-Cox transformation, we use the A-class inventory plots of Miyun County Seventh National forest resources, the same period TM image data through geometric correction and 30 m resolution DEM data as basic data, to classify and estimate forest volume of Miyun County in Beijing, and to correct the stock volume by Box-Cox transformation. The analysis results are as follows: Stock volume estimation by classification can effectively improve the estimation accuracy and the ability of model forecast and classification modeling can be effectively used into stock volume estimation; Under the same conditions of factors and model sample plots, although the accuracy is slightly reduced after correcting the model by Box-Cox correction of stock volume, it can greatly improve the prediction ability and applicability of the model. Stock volume estimation needs regression diagnostics. When the abnormal plots are more and they affect Gauss-Markov assumptions, we can introduce Box-Cox transformation.