在大量的植被冠层遥感模型中,辐射度模型作为一种计算机模拟模型具有很多优点,它对理解植被-辐射相互作用过程和研究植被冠层辐射机理具有重要的理论价值。传统的辐射度模型假设冠层内叶片为朗伯体。但随着对叶片特性的深入研究发现,叶片的反射特性不能单纯的看成是理想漫反射过程。为了消除传统辐射度模型的朗伯假设,完善辐射度模型,本文利用Phong光照模型模拟叶片表面非朗伯(镜面反射)部分的分布特性,在基于真实结构冠层场景的辐射度模型(RGM)的基础上,增加了叶片镜面反射分量的计算。文中以玉米冠层为例,比较并分析了叶片的镜面反射分量对冠层辐射分布的影响。最后,从辐射度模型原理出发,进一步推导得出了计算冠层镜面反射分量的辐射度公式,从理论上证明了在可见光-近红外区的冠层镜面反射分量与波长无关。本文使用的方法具有简单、易于实现的特点。经过扩展后的辐射度模型既保留了传统辐射度模型的优点,又增加了该模型的功能和模拟范围。
As one of the computer simulation models, the Radiosity-Graphics combined Model(RGM) has many advantages in calculating the bidirectional reflectance factor (BRF). Because it takes advantage of radiosity theory and computer graphics technique, the model can contain much more detailed and complex structures of vegetation canopy and take reflection, transmission and multiple scattering into account, which is useful to understand the interaction between the light and the canopy. A hypothesis of Lambertian is made in the general radiosity theory, namely, all surfaces of components (i. e.., leaves, stem and background) in the scene are Lambert reflection/transmission. In fact, studies of the properties of leaves have shown that the bidirectional reflectance distribution functions(BRDF) of most leaves' surfaces are not isotropic. In order to apply RGM to calculate the radiance distribution caused by the non-Lambert(specular) component, a semi- experimental Phong model is used to evaluate the specular radiosity from leaves' surfaces. This method is applied to the maize canopy, and the results are analyzed. As an interesting experiment, this extended RGM, which includes diffuse and specular component at the same time, not only keeps the advantages of the general radiosity theory, but also eliminates the hypothesis of Lambertian in vegetation scene.