Synthetic aperture radar (SAR) features have been demonstrated that they have the potentiality to improve forest above ground biomass (AGB) estimation accuracy, especially including polarimetric information. Genetic algorithms (GAs) have been successfully implemented in optimal feature identification, while support vector regression (SVR) has great robustness in parameter estimation. The use of combined GAs and SVR can improve the accuracy of forest AGB estimation through simultaneously identifying the optimal SAR features and selecting the SVR model parameters. In this article, 14 SAR polarimetric features were extracted from C-band and L-band full-polarization SAR images and worked as input SAR features, respectively. C-band data was acquired on GaoFen-3 mission, we also call it GF-3 image. L-band data was ALOS-2 PALSAR-2 data. Both feature subsets from GF-3 and ALOS-2 PALSAR-2 and SVR hyper parameters used in the forest AGB estimation were optimized by a GA processing, where 8 different settings of 3 kinds of parameters, as 512 kind of different combinations were applied for SVR hyper parameters searching field. The results of GA-SVR performance using the two datasets were presented and compared with two traditional methods: the algorithm of GA feature selection companied with default SVR parameters (GA+ default SVR), and the algorithm of GA feature selection companied with grid searching for SVR parameter selection (GA+Grid SVR). The results showed that the proposed GA-SVR algorithm improved the forest AGB estimation accuracy with cross-validation coefficient of 80.21% for GF-3 and 71.41% for ALOS-2 PALSAR-2 data.

中文翻译：

---

利用 SAR 数据改进森林地上生物量估算的 GA-SVR 算法


合成孔径雷达（SAR）功能已被证明具有提高森林地上生物量（AGB）估计精度的潜力，特别是包括偏振信息。遗传算法（GA）已成功应用于最优特征识别，而支持向量回归（SVR）在参数估计中具有很强的鲁棒性。结合使用GA和SVR可以通过同时识别最佳SAR特征和选择SVR模型参数来提高森林AGB估计的准确性。本文从C波段和L波段全偏振SAR图像中提取了14个SAR偏振特征，分别作为输入SAR特征。 C波段数据是在高分三号任务中获取的，我们也称其为GF-3图像。 L波段数据是ALOS-2 PALSAR-2数据。森林 AGB 估计中使用的 GF-3 和 ALOS-2 PALSAR-2 和 SVR 超参数的两个特征子集均通过 GA 处理进行优化，其中 3 种参数的 8 种不同设置，应用了 512 种不同的组合SVR超参数搜索字段。给出了使用两个数据集的 GA-SVR 性能结果，并与两种传统方法进行了比较：带有默认 SVR 参数的 GA 特征选择算法（GA+默认 SVR）和带有网格搜索 SVR 的 GA 特征选择算法参数选择（GA+Grid SVR）。结果表明，所提出的GA-SVR算法提高了森林AGB估计精度，GF-3数据的交叉验证系数为80.21%，ALOS-2 PALSAR-2数据的交叉验证系数为71.41%。