极化合成孔径雷达（PolSAR）能够实现全天时、全天候的成像，因此该数据成为遥感数据的主要来源之一。其中地物分类是极化SAR数据解译的重要研究内容，已成为该研究领域的热点之一，目前在军事和民事领域都有着广泛的应用。近年来深度学习已在多个研究领域取得了显著成果，并且在极化SAR图像解译领域也获得了一定的成效。与传统的图像分类方法相比，深度学习方法具有自动提取特征、泛化性能强及获取较高准确率等优势。围绕极化SAR数据解译中的地物分类问题，对现有采用深度学习的极化SAR图像地物分类方法进行综述。根据深度学习中不同的网络模型，主要从三方面对极化SAR地物分类研究进行了详细叙述，即基于深度信念网络，稀疏自编码网络以及卷积神经网络的图像分类模型。最后，通过与经典的极化SAR分类方法进行性能评估和比较，总结采用深度学习的极化SAR地物分类方法的优势与不足，同时对该领域未来的发展趋势进行分析和探讨。

As one of the important research contents of Synthetic Aperture Radar(SAR) image interpretation, Polarimetric Synthetic Aperture Radar(PolSAR) terrain classification has attracted more and more attention from scholars at home and abroad. Different from natural images, the PolSAR dataset not only has unique data attributes but also belongs to a small sample dataset. Therefore, how to make full use of the data characteristics and label samples is a key consideration. Based on the above problems, a new network on the basis of UNet for PolSAR terrain classification-Multiscale Separable Residual Unet(MSR-Unet) is proposed in this paper. In order to extract separately the spatial and channel features of the input data while reducing the redundancy of features, the ordinary 2D convolution is replaced by the depthwise separable convolution in MSR-Unet. Then, an improved multi-scale residual structure based on the residual structure is proposed. This structure obtains features of different scales by setting convolution kernels of different sizes, and at the same time the feature is reused by dense connection, using the structure can not only increase the depth of the network to a certain extent and obtain better features, but also enable the network to make full use of label samples and enhance the transmission efficiency of features information, thereby improving the classification accuracy of PolSAR terrain. The experimental results on three standard datasets show that compared with the traditional classification methods and other mainstream deep learning network models such as UNet, the MSRUnet can improve average accuracy, overall accuracy and Kappa coefficient in different degrees and has better robustness.

为了探究独流减河入海口微塑料空间分布特征,本研究于2021年3月采集了该区域不同环境介相样品,通过消解浮选和激光红外成像系统对样品中微塑料种类、丰度、粒径和形状进行了深入研究.结果显示,在水样中检测到的微塑料的浓度范围为227～402 items/L.共检测到17种微塑料,其中丰度最高的是聚丙烯(PP)、聚乙烯(PE)、聚酰胺(PA)、聚对苯二甲酸乙二醇酯(PET)和聚甲基丙烯酸甲酯(PMMA);沉积物中微塑料数量丰度为8 430～12 600 items/kg,共检测到18种微塑料,其中丰度最高的是PE、PP、PET、PA和聚氯乙烯(PVC).在水样和沉积物中,粒径分布表现为随着粒径减小微塑料的数量急剧增加;在形状分布上,除少部分大粒径的PET、PA和PP表现为纤维外,绝大多数微塑料为颗粒态.本研究表明,微塑料随独流减河从淡水到海水的传输过程中,随着盐度和密度提高,水体中微塑料丰度显著提高;而同样由于密度的差异及使用用途的差异,导致水样和沉积物之间微塑料组成结构具有明显差异.揭示了水体及微塑料密度及微塑料用途是影响河口区域微塑料空间分布的重要因素.

The structural design of constrained joints is an effective way to realize the type synthesis of constrained metamorphic mechanism. In previous studies, constrained joints are often represented in the form of kinematic diagrams of mechanism, which doesn’t provide enough reference to the practical application of metamorphic mechanisms. On the basis of summarizing the constraint forms and resistance characteristics of the commonly used variable-constraint constrained joints, and according to the force analysis of the augmented Assur group with constrained joints, a new constrained revolute joint with combined variable-constraint of spring force and geometric constraint is designed, which has advantages such as simple structure, reliable working ability and passive adaptive capability to operational conditions. Based on this constrained joint, a two-configuration constrained metamorphic mechanism with under-actuated is designed, the force analysis of the mechanism is completed, and the effectiveness and feasibility of design approach proposed in this paper is verified for the constraint parameters of the invented metamorphic joint. The novel combinatorial variable-stiffness constrained joint can obtain stable and reliable configuration switching ability, which provides a feasible way for promoting the engineering application of metamorphic mechanism.

构态切换过程的力学调控机理和构态切换能力的定义及评价是变胞机构实现构态平稳切换的理论基础.考虑工程实际中的误差随机性问题,建立构态切换能力的概率评估模型是降低变胞机构运动功能失效概率的关键.基于多构态欠驱动约束变胞源机构的结构组成原理建立了含变胞副的扩展Assur杆组模块化受力分析模型.根据变胞副的等效阻力梯度模型,完成了确定性条件下约束变胞机构构态切换的力学机理建模.进而,考虑系统内外部误差的随机性,建立了约束变胞机构构态切换时刻的可靠性评估模型.考虑构态切换相邻时刻运动功能失效的影响,利用系统可靠性理论提出了约束变胞机构构态切换的区间可靠度计算模型.最后,以折纸变胞机构为例,验证了提出的构态切换能力概率评估方法的可行性和有效性.为开展以获得高概率的构态切换重复执行能力为目标的约束变胞机构可靠性优化设计奠定了基础,同时对推进变胞机构的工程应用具有理论和实际意义.

According to the small sample characteristics of polarimetric synthetic aperture radar (PolSAR) data and its unique data attributes, a new network architecture for PolSAR images classification based on Unet is proposed in this paper. Fully considering the characteristics of PolSAR data, the spatial features and channel features of the input data are extracted respectively by the depthwise separable convolution and avoid extracting redundant features. In order to improve the classification accuracy, the residual structure is used to increase the depth of the network and fully transmit the characteristics information of PolSAR data. The experimental results clearly demonstrate that the architecture we proposed can achieve better classification accuracy than other PolSAR images classification methods.

When we use deep learning to classify PolSAR images, the lack of labeled samples will affect the classification performance. This paper uses the ACGAN model to expand the training samples of PolSAR data, utilizing the generated samples and original samples together to train CNN for PolSAR image classification. In addition to random noise, the ACGAN model also inputs additional relevant guidance information to ensure that the new generated samples are more similar to the original data. Compared with GAN, the discriminator of ACGAN can distinguish not only whether the data is real or not, but also the class label of the data. Subsequently, the validity of our proposed method is verified on the San Francisco data set. Compared with other classical PolSAR classification methods, the accuracy of our proposed method is improved.