基于广义逆矩阵的BRDF模型参数拟合方法

Parameter fitting method of BRDF model based on generalized inverse matrix

为了提高二向性反射率分布函数(BRDF)模型参数拟合的效率,充分了解地物目标的光谱方向特性,提出了一种基于广义逆矩阵的BRDF模型参数拟合方法。首先分析了地物光谱反射的方向特性,然后使用成像光谱仪测量不同方向条件下目标的光谱反射率,将实验条件与先验测量数据代入BRDF模型中,建立广义逆矩阵等式并拟合出测量目标的模型参数;利用拟合的目标BRDF模型参数反推任意特定方向条件下的光谱反射率,并与实测数据作对比,采用光谱角制图法度量其相似度。结果表明,草地和迷彩雨衣的拟合光谱曲线与实测光谱曲线的相似度很高,分别达到了0.1307和0.0896,证实了该参数拟合方法的有效性,同时也验证了BRDF模型的泛化拟合能力。该方法原理简单、快速有效,可为其它类型的BRDF模型参数拟合与后续地物目标光谱特性研究提供参考。

In order to improve the efficiency of parameter fitting of binary reflectance distribution function(BRDF)model and understand the spectral direction characteristics of ground object more specifically,a method of parameter fitting of BRDF model based on generalized inverse matrix was proposed.Firstly,the directional characteristics of spectral reflection of ground objects were analyzed,and then the spectral reflectance of targets in different directions was measured by imaging spectrometer.The experimental conditions and a priori measurement data were substituted into BRDF model,the generalized inverse matrix equation was established,and the model parameters of measurement targets were fitted.Using the BRDF model parameters of fitting the push any particular direction under the condition of spectral reflectance and compared with experimental data.Spectral angle mapping method was used to measure its similarity and the results showed that the grass and camouflage raincoat fitting spectrum curve and the measured spectral curve similarity is very high,up to the 0.1307 and 0.0896,confirmed the validity of the parameter fitting method,the ability to verify the generalization of the BRDF model fitting.This method is simple in principle,fast and effective,and provides reference for other types of BRDF model parameter fitting and subsequent research on spectral characteristics of ground objects.