【作者】 赵芳；

【导师】 冯仲科；

【作者基本信息】 北京林业大学 ， 林业装备工程， 2014， 博士

【摘要】 森林资源调查与监测是森林管理与经营的基础,查清、落实好森林资源质量和数量可为森林资源的可持续发展提供有力依据。经过了半个世纪的发展,我国的森林资源调查与监测已经形成了一套符合我国国情、符合我国林业发展特点的较为完整的森林调查体系。目前,如何利用先进的科学手段获取森林资源是林业工作者们的研究方向,并对森林的变化规律、变化状况进行监测,从而更加全面、客观的管理与经营森林信息。本研究尝试着对森林资源调查信息获取的传统方法的改变,分别利用近景摄影测量技术、航空摄影测量技术及遥感反演技术对测树因子信息进行提取,立足于降低劳动强度,提高森林信息的获取的效率,使森林资源调查做到自动化、智能化、现代化、科学化,为实现做到“数字森林”提供最基础的数据保障,为全面实现数字化健康森林体系做好有力基石。文章主要研究内容及结论如下：1、本研究以近景摄影测量理论与技术为基础,分别利用普通数码照相机和3D数码照相机对单株立木进行拍照,结合地面摄影测量理论和方法对单株立木求解其胸径值：(1)利用数字图像量测,运用数字图像比例关系来解算胸径,估测精度为87.65%。(2)通过经典的共线方程建立像方坐标与物方三维空间坐标的关系式来解算拍照的单株立木的胸径值,估测精度为88.87%。(3)利用3D数码照相机能同时拍摄到一个像对的功能,采用水平正直摄影的方式进行拍摄,对同时拍摄到的像对照片解算胸径,虽然结果效果不佳,但是为利用3D相机求解树木参数提供一种思路。2、以航空摄影测量的理论和方法为基础,首先对森林资源调查的各项测树因子进行提取,通过不同的提取方法对单株尺度的树高、冠幅和林分尺度的林分平均高、林分平均冠幅、郁闭度等进行了估测,并利用该林分因子建立了航空蓄积量模型。然后对利用航空摄影测量的方法所获取的各项测树因子建立模型,以估测单株立木胸径,最后进行精度验证,比较各个模型进度,其中以树高、冠幅来推算胸径的估测模型精度最佳。3、通过卫星遥感技术分别利用资源三号卫星影像和Landsat TM卫星影像估算生物量。其中不仅将光谱因子、植被因子和地形因子作为反演生物量的因子,还添加了纹理因子共同参与反演,同时,各个像元值采用平均中心像元法(3×3尺度),作为该中心像元的特征值。最后通过多元线性回归,得出最优森林生物量模型,并以此估算出北京市森林生物量及生物量分布图。通过研究获取测树因子的三种不同尺度,从局部到整体,对森林在不同空间范围内的各个森林调查因子的采集进行研究,这一研究对当前的森林资源管理具有较大的现实意义。文章所提出的主要技术对改变传统森林资源调查费时、费力的现状,提高森林资源调查获取数据的水平、降低野外调查劳动强度和生产成本等方面具有重要意.义。更多还原

【Abstract】 Forest resources survey and monitoring is the foundation of forest management and operation, which can provide a strong basis for forest resources sustainable development depends on identify and implement good quality of forest resources. After half a century development, China’s forest resources survey and monitoring has formed a more comprehensive forest inventory system, according with China’s national conditions and forestry developments. At present, directions of the forestry workers is to obtain the forest resources using the advanced scientific, besides, to monitor forest variation situation changes more comprehensive, objective information for forest management.This study attempts to change the traditional methods of forest resources survey, respectively using close-range photogrammetry, aerial photography and remote sensing measurement techniques to obtain forest management information, which based on reducing labor intensity and improving forest information achieve efficiency, made the forest resources survey automatic, intelligent, modern and scientific Provide the best basis for the realization of a "digital forest" and lay the cornerstone for the digital healthy forest system.1. This study is based on the close-range photogrammetry theory and technology, respectively using an ordinary digital camera and a3D digital camera to take pictures for the plant, combined with terrestrial photogrammetry theory and methods for solving its DBH plant stumpage value:Using digital images measurements and proportional to DBH solver, estimation accuracy achieved87.65%.Establishing the relationship between the image of coordinates and the type of the object in three-dimensional space coordinates of the single tree to calculate the value of the picture through diameter at breast height by classical collinearity equation, estimation accuracy was88.87%. Using the function of the3D digital camera to get a group image pair at the same time, as to adopt the way of horizontal straight photography, the photos taken at the same time as the calculating DBH was ineffective, but for the use of3d camera will provide an idea to solve the parameters of trees.2.Collected various measurement factors of the tree at first and estimated the plant scale of tree height, crown width and the stand scale average height, average Stand Crown, crown density, which is based on the methodology theory and aerial photogrammetry of forest resource survey. Establishment the air volume model used the stand factors. Then the method of the aerial photography acquired factor of tree measuring model, to estimate individual tree diameter at breast height, finally, to verify the accuracy of each model, comparison of progress, the tree height, diameter at breast height estimation model to calculate the optimal precision.3. Biomass estimated through satellite remote sensing technique using resources three satellite images and Landsat TM images. Not only the spectroscopic, vegetation and terrain factor as inversion biomass factor, but also added texture factor involved in inversion, at the same time, each pixel value used the average center pixel (3×3scale), as the center pixel value. Finally, through multiple linear regression, achieved the biological optimal forest model, and estimated the forest biomass and biomass distribution map of Beijing city.Research and measurement of three different scale factor of tree, from local to overall, to research the forest investigation factor of forest in different space within the scope of the collection has important practical significance to the management of forest resource. The main technology that proposed in this paper is to change the traditional forest resource survey time-consuming, laborious, improve the level of survey data, reduce the important field survey labor intensity and production cost of forest resources.更多还原