覆盖度（Fractional Vegetation Cover，FVC）和叶面积指数（Leaf Area Index，LAI）是衡量地表植被生长状况的重要植被结构参数，可为植被冠层表面的物质能量交换提供定量化信息，是植被遥感中最为重要的反演结构参数。地面实测值既是遥感定量反演植被参数的先验知识，又是评价遥感反演精度的重要标准。FVC、LAI的实地测量方法可分为直接测量法和间接测量法。前者往往需要较大的工作量、复杂的测量过程和破坏性采样测量，因此在实地测量中多使用仪器进行间接测量的方法来获取FVC、LAI，如数码相机照相法测FVC，LAI-2000、TRAC测LAI等。然而目前对间接测量方法精度分析的研究较少，究其原因是实地测量中难以获取FVC、LAI真值。与此同时，计算机真实结构模拟模型可获取模拟植被场景的FVC、LAI真值。基于此，本文利用计算机模拟的方法对数码相机照相法测FVC，多光谱冠层成像仪（Multispectral Canopy Imager，MCI）测LAI的测量精度进行了研究。本文利用实地测量和模拟产生的植被冠层场景结构数据，采用基于物理模型的光线追踪法（Physically Based Ray-Tracing，PBRT）模拟光与冠层组分的相互作用。由于照相法测FVC、MCI测LAI都采用数码相机对场景成像的测量方式，所以在模型中模拟了一相机对场景进行拍摄，根据照相法测FVC和MCI测LAI的测量过程设置模拟相机参数，获取模拟照片，再分别按照FVC和LAI的计算方法从模拟的照片中提取FVC、LAI值。通过对比模拟场景的FVC、LAI真值，分析仪器测量精度。对于照相法测FVC的精度分析，本文利用模拟的随机分布场景与L系统生成的玉米场景分析了测量过程中相机拍摄高度、非垂直拍摄及拍摄时光照情况对FVC计算结果的影响；利用Onyx Tree软件生成的梨树林场景分析了照相法测FVC过程中采样策略对测量结果的影响。此外，本文对MCI的测量过程进行了模拟，分析了MCI的测量精度。对于MCI测量LAI的精度分析，本研究利用生成的梨树林场景，采用有限长度平均法和间隙大小分布法计算了模拟场景聚集指数、有效叶面积指数和木质组分面积指数在不同天顶角、方位角的分布。并将LAI计算值与场景真值进行对比，分析了MCI测量森林LAI的精度。

Plant parameters such as Fractional Vegetation Cover (FVC) and Leaf Area Index (LAI) are important indicators of vegetation growth conditions. They can provide quantitative material and energy exchange information for the vegetation crown and they are the most important structural parameters that remote sensing-retrieved.The data measured in situ are not only the evaluation criteria for remote sensing-retrieved plant parameters, but also the foundation for developing quantitative vegetation models in remote sensing. Field measurements can be classified into direct methods and indirect methods. Due to the numerous complex manual measurements and destructive sampling of direct methods, the indirect methods are now used widely to obtain the plant parameters. For example, digital camera is used to measure the FVC and LAI-2000, TRAC are used to measure the LAI. However, there’re few studies on the accuracy evaluation of these indirect methods. This is mainly because it’s difficult to obtain the true values of the plant parameters in field measurements. Meanwhile, for the virtual canopy scenes in computer, it’s easy to obtain the accurate plant parameters values of simulated plant scene. Thus, this thesis uses computer simulation method to evaluate the accuracy of the indirect methods of digital photography and Multispectral Canopy Imager (MCI) which is an instrument developed by Yan et.al (2008) to measure the clumping index, effective LAI and Wood Area Index (WAI) at different viewing zenith angle and azimuth angle.Based on the measured and simulated plant structural data, the Physical Based Ray-Tracing (PBRT) algorithm was used to calculate the interactions of light and plant crown. A virtual digital camera was also simulated to get the simulated images for both digital photography and MCI just as using the digital camera to measure the FVC and LAI. We set the digital camera just as we did in the field measurements to get the simulated images. Then the FVC and LAI were calculated from the simulated images with the same indirect methods used in digital photography and MCI. By comparing the true plant parameters values with calculated ones, we can evaluate the accuracy of these indirect methods.As for digital photography, a simulated random distributed scene and a maize scene generated by the L-system were used to evaluate the influence of the shooting height, viewing zenith angle of digital camera and shooting time. Besides, a pyrus scene generated by Onyx software was used to evaluate the influence of the sampling strategy when measuring FVC.The measuring process of MCI was also simulated to evaluate the accuracy of MCI. Same algorithms in MCI were used to calculate the simulated images to extract the clump index, woody area index, effective leaf area index at different viewing zenith angle and azimuth angle of the simulated pyrus scene. The accuracy of MCI was evaluated by comparing the true LAI values of the simulated scene with calculated LAI values.