[论文解读] Classification of polarimetric SAR images using compact convolutional neural networks

Mete Ahishali, Serkan Kıranyaz+2 more|arXiv (Cornell University)|Apr 26, 2022

Synthetic Aperture Radar (SAR) Applications and Techniques参考文献 47被引用 24

一句话总结

本文提出一种紧凑且自适应的卷积神经网络（CNN）用于全息SAR（PolSAR）图像分类，采用滑动窗口方法，避免了人工特征提取。该方法在极小的训练数据量（<0.1%）下实现高达99.39%的准确率，计算成本低，可在标准硬件上实现实时处理，即使在小窗口尺寸（7×7至19×19）下亦可实现。

ABSTRACT

Classification of polarimetric synthetic aperture radar (PolSAR) images is an active research area with a major role in environmental applications. The traditional Machine Learning (ML) methods proposed in this domain generally focus on utilizing highly discriminative features to improve the classification performance, but this task is complicated by the well-known ?curse of dimensionality? phenomena. Other approaches based on deep Convolutional Neural Networks (CNNs) have certain limitations and drawbacks, such as high computational complexity, an unfeasibly large training set with ground-truth labels, and special hardware requirements. In this work, to address the limitations of traditional ML and deep CNN-based methods, a novel and systematic classification framework is proposed for the classification of PolSAR images, based on a compact and adaptive implementation of CNNs using a sliding-window classification approach. The proposed approach has three advantages. First, there is no requirement for an extensive feature extraction process. Second, it is computationally efficient due to utilized compact configurations. In particular, the proposed compact and adaptive CNN model is designed to achieve the maximum classification accuracy with minimum training and computational complexity. This is of considerable importance considering the high costs involved in labeling in PolSAR classification. Finally, the proposed approach can perform classification using smaller window sizes than deep CNNs. Experimental evaluations have been performed over the most commonly used four benchmark PolSAR images: AIRSAR L-Band and RADARSAT-2-C-Band data of San Francisco Bay and Flevoland areas. Accordingly, the best obtained overall accuracies range between 92.33-99.39% for these benchmark study sites.

研究动机与目标

• 解决深度CNN在PolSAR分类中计算复杂度高和数据需求大的问题。
• 克服传统机器学习方法依赖人工高维特征工程的局限性。
• 在极小标注训练数据和低计算开销下实现高精度的PolSAR分类。
• 在不同PolSAR数据集和窗口尺寸下展示模型的鲁棒性与泛化能力。
• 通过消除对专用硬件和复杂预处理的依赖，实现真正的实时处理。

提出的方法

• 提出一种紧凑、自适应的四层CNN架构，采用参数共享和参数减少策略，以最小化计算负载。
• 采用滑动窗口分类方法，直接从二阶PolSAR描述符处理小图像块（7×7至19×19像素）。
• 直接处理原始复数后向散射矩阵，无需人工特征提取，利用端到端学习实现特征提取与分类的联合优化。
• 采用多 look 相干矩阵和极化描述符（如SPAN、Pauli分解）作为输入通道（3–6通道），以保留物理散射信息。
• 采用自适应学习率的反向传播算法，在有限标注数据下高效训练网络。
• 通过超参数调优（神经元和层数的乘数）评估模型鲁棒性并避免过拟合。

实验结果

研究问题

• RQ1 紧凑CNN能否在极小标注训练数据下实现PolSAR图像的高分类准确率？
• RQ2 与传统机器学习方法和深度CNN相比，该方法在计算效率和数据需求方面表现如何？
• RQ3 使用紧凑CNN进行准确PolSAR分类的最优窗口尺寸是什么？
• RQ4 该模型能否在不同PolSAR数据集（如旧金山湾、弗莱沃兰）之间实现泛化，通过跨站点验证？
• RQ5 模型性能对网络深度和宽度（超参数）变化的敏感性如何？

主要发现

• 所提出的紧凑CNN在四个基准PolSAR数据集（旧金山湾和弗莱沃兰的AIRSAR L波段与RADARSAT-2 C波段）上实现了92.33%至99.39%的整体准确率。
• 在SFBay_L数据集上，采用四通道输入和21×21窗口尺寸时，取得99.39%的整体准确率。
• 在跨站点验证中，模型泛化能力良好：当在Flevo_C上训练并在Flevo_L上测试时，整体准确率达到88.11%，即使合并了作物类别。
• 该方法的训练数据量不足总数据的0.1%，显著降低了标注成本和数据依赖性。
• 增加网络深度或宽度（如m=8, n=1或n=2）导致过拟合和不收敛，表明紧凑性对模型稳定性至关重要。
• 模型在7×7至19×19的窗口尺寸范围内均保持高准确率，证明其对小图像块的鲁棒性。