An optimized Back-Propagation network (BP network) based on Genetic Algorithm (GA) was introduced to construct bidirectional reflectance distribution function (BRDF) model. To verify the performance of GA-BP network, two different kinds of space target materials were used for experiment. Based on the experimental data, we used GA to simulate the undetermined parameters of a five-parameter BRDF model, and used GA-BP network and BP network to construct a new BRDF model respectively. The fitting results manifest that the GA-BP network is suitable for construct a new BRDF model and outperforms the five-parameter BRDF model in speed and accuracy under the same condition.

中文翻译：

---

基于GA-BP网络的可见光谱空间目标双向反射分布函数模型

引入基于遗传算法(GA)的优化反向传播网络(BP网络)构建双向反射分布函数(BRDF)模型。为了验证 GA-BP 网络的性能，使用两种不同的空间目标材料进行实验。基于实验数据，我们使用GA模拟了一个五参数BRDF模型的待定参数，并分别使用GA-BP网络和BP网络构建了一个新的BRDF模型。拟合结果表明，GA-BP 网络适合构建新的 BRDF 模型，在相同条件下的速度和精度上均优于五参数 BRDF 模型。