基于特征融合的全卷积网络极化SAR分类方法

Fully convolutional network PolSAR classification based on features fusion

极化合成孔径雷达可以工作在多个极化方式下,综合利用多种极化回波数据实现地物分类是极化数据处理的一个重要应用。目前将卷积神经网络应用于极化地物分类领域仍存在相应问题,包括多维极化分解特征信息给网络带来的信息冗余与维数灾难,逐像素切片预测导致分类效率低下。针对以上问题,提出了一种基于特征融合的全卷积网络模型。首先,设计两路编码层分支的全卷积网络结构,分别针对极化分解特征与极化散射特征提取深层特征,实现多维特征信息分离。然后采用注意力特征融合机制实现两路分支的特征融合,通过共享连接层学习通道注意力权值,重新分配网络的学习能力。此外引入改进的空洞空间金字塔模块,以提升模型的多尺度预测能力。实验结果表明,算法在两个不同地区的极化数据集的总体分类精度分别达到96.43%与99.60%,预测耗时分别为17.3 s与10.1 s。在不显著增加预测耗时的同时提升了分类精度,验证了算法的有效性。

Polarized synthetic aperture radar can work in multiple polarization modes.Using multiple polarization echo data to achieve ground feature classification is an important application of polarization data processing.At present,there are still some problems in the application of convolutional neural network in the field of polarization feature classification.Including the information redundancy and dimension disaster caused by multi-dimensional polarization decomposition feature information.To mitigate these problems,proposes a fully convolutional network model based on feature fusion.Firstly,by designing a full convolutional network structure with two encoding layer branches,the deep features are extracted for the polarization decomposition feature and polarization scattering feature respectively to realize the separation of multi-dimensional feature information.Then,the attention feature fusion mechanism is adopted to realize the feature fusion of two branches,and the learning ability of the network is redistributed by sharing the attention weight of the connection layer learning channel.In addition,an improved atrous space pyramid pooling is introduced to improve the multi-scale prediction ability of the model.The experimental results show that the overall accuracy of polarization data sets in two different regions is 96.43% and 99.60% respectively,and the prediction time is 17.3 s and 10.1 s.The classification accuracy is improved without greatly increasing the prediction time,and the effectiveness of the algorithm is verified.