Isotropization of Quaternion-Neural-Network-Based PolSAR Adaptive Land Classification in Poincare-Sphere Parameter Space

Kinugawa, Kazutaka; Shang, Fang; Usami, Naoto; Hirose, Akira

doi:10.1109/lgrs.2018.2831215

Cited by 34 publications

(14 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, the common disadvantages of these methods are usually complicated parameter estimation and limited model applicability [27,28]. Recently, some classifiers with the training-testing format, which work directly on the PolSAR CV data, have constituted an active area of research [2,20,[29][30][31][32][33][34]. Among them, complex-valued networks provide results comparable to networks designed for real-valued input [32][33][34].…”

Section: Introductionmentioning

confidence: 99%

“…In addition, as we know, due to the imaging mechanism, PolSAR images are heavily contaminated by the inherent speckles [1]. Note that some of the aforementioned methods based on PolSAR matrices only consider the polarimetric characteristics [20,27,29,30]. The existence of speckles may make classification results include many misclassified pixels and degrade the quality of classification, especially when the training samples are limited [37,38].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

PolSAR Image Classification Using a Superpixel-Based Composite Kernel and Elastic Net

Cao

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

The presence of speckles and the absence of discriminative features make it difficult for the pixel-level polarimetric synthetic aperture radar (PolSAR) image classification to achieve more accurate and coherent interpretation results, especially in the case of limited available training samples. To this end, this paper presents a composite kernel-based elastic net classifier (CK-ENC) for better PolSAR image classification. First, based on superpixel segmentation of different scales, three types of features are extracted to consider more discriminative information, thereby effectively suppressing the interference of speckles and achieving better target contour preservation. Then, a composite kernel (CK) is constructed to map these features and effectively implement feature fusion under the kernel framework. The CK exploits the correlation and diversity between different features to improve the representation and discrimination capabilities of features. Finally, an ENC integrated with CK (CK-ENC) is proposed to achieve better PolSAR image classification performance with limited training samples. Experimental results on airborne and spaceborne PolSAR datasets demonstrate that the proposed CK-ENC can achieve better visual coherence and yield higher classification accuracies than other state-of-art methods, especially in the case of limited training samples.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

PolSAR Image Classification Using a Superpixel-Based Composite Kernel and Elastic Net

Cao

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Quaternion neural networks (QNNs) [17] have been employed for PolSAR land classification [18]- [20]. In QNN methods, a three dimensional vector is regarded as a single quaternion.…”

Section: Introductionmentioning

confidence: 99%

Full-learning quaternion convolutional neural networks and cross-entropy-based confluence of differently represented data for PolSAR land classification

Matsumoto¹

2021

Preprint

View full text Add to dashboard Cite

ALOS2 Dataset1<div><div>Odi_SceneId="SARD000000088012-00036-019-001"</div><div>Odi_SiteDateTime="PROCESS:JAPAN-JAXA-ALOS2-EICS 20151207 072121"</div><div>Scs_SceneID="ALOS2066310850-150815"</div></div><div><br></div><div>ALOS2 Dataset2</div><div><div>Odi_SceneId="SARD000000088012-00036-044-002"</div><div>Odi_SiteDateTime="PROCESS:JAPAN-JAXA-ALOS2-EICS 20210113 054828"</div><div>Scs_SceneID="ALOS2066310860-150815"</div></div>

show abstract

“…Shang et al [32] suggested a complexvalued feedforward neural networks in the Poincare sphere parameter space. Moreover, an improved quaternion neural network [33] and a quaternion autoencoder [34] have been proposed for PolSAR land classification. Recently, a complexvalued CNN (CV-CNN) specifically designed for PolSAR image classification has been proposed by Zhang et al [35], where the authors derived a complex backpropagation algorithm based on stochastic gradient descent for CV-CNN training.…”

Section: Introductionmentioning

confidence: 99%

Pixel-Wise PolSAR Image Classification via a Novel Complex-Valued Deep Fully Convolutional Network

Cao

Zhang

et al. 2019

Remote Sensing

View full text Add to dashboard Cite

Although complex-valued (CV) neural networks have shown better classification results compared to their realvalued (RV) counterparts for polarimetric synthetic aperture radar (PolSAR) classification, the extension of pixel-level RV networks to the complex domain has not yet thoroughly examined. This paper presents a novel complex-valued deep fully convolutional neural network (CV-FCN) designed for PolSAR image classification. Specifically, CV-FCN uses PolSAR CV data that includes the phase information and utilizes the deep FCN architecture that performs pixel-level labeling. It integrates the feature extraction module and the classification module in a united framework. Technically, for the particularity of PolSAR data, a dedicated complex-valued weight initialization scheme is defined to initialize CV-FCN. It considers the distribution of polarization data to conduct CV-FCN training from scratch in an efficient and fast manner. CV-FCN employs a complex downsampling-then-upsampling scheme to extract dense features. To enrich discriminative information, multi-level CV features that retain more polarization information are extracted via the complex downsampling scheme. Then, a complex upsampling scheme is proposed to predict dense CV labeling. It employs complex max-unpooling layers to greatly capture more spatial information for better robustness to speckle noise. In addition, to achieve faster convergence and obtain more precise classification results, a novel average cross-entropy loss function is derived for CV-FCN optimization. Experiments on real PolSAR datasets demonstrate that CV-FCN achieves better classification performance than other state-of-art methods.Index Terms-Complex-valued deep fully convolutional neural network (CV-FCN), polarimetric synthetic aperture radar (Pol-SAR) image classification, pixel-level labeling.

show abstract

Isotropization of Quaternion-Neural-Network-Based PolSAR Adaptive Land Classification in Poincare-Sphere Parameter Space

Cited by 34 publications

References 9 publications

PolSAR Image Classification Using a Superpixel-Based Composite Kernel and Elastic Net

PolSAR Image Classification Using a Superpixel-Based Composite Kernel and Elastic Net

Full-learning quaternion convolutional neural networks and cross-entropy-based confluence of differently represented data for PolSAR land classification

Pixel-Wise PolSAR Image Classification via a Novel Complex-Valued Deep Fully Convolutional Network

Contact Info

Product

Resources

About