Hyperspectral Anomaly Detection With Attribute and Edge-Preserving Filters

Kang, Xudong; Zhang, Xiangping; Li, Shutao; Li, Kenli; Li, Jun; Benediktsson, Jón Atli

doi:10.1109/tgrs.2017.2710145

Cited by 358 publications

(157 citation statements)

References 37 publications

Supporting

Mentioning

155

Contrasting

Unclassified

Order By: Relevance

“…It has the spectral range of 370 nm to 2510 nm. After removing some noisy bands [53], totally 207 spectral channels are remained in the experiment. This image has the size of 100 × 100.…”

Section: Data Setsmentioning

confidence: 99%

“…The difference is that this image has a higher spatial resolution of about 3.5 m for it is acquired from a low-altitude aircraft through scanning the Gainesville urban scene. Considering the effects of some noisy bands, we keep 191 spectral bands, after removing 33 useless bands referring to the work [53]. There are 100 × 100 pixels in the image, where totally 11 abnormal targets are located in.…”

Section: Data Setsmentioning

confidence: 99%

See 1 more Smart Citation

Hyperspectral Anomaly Detection Based on Separability-Aware Sample Cascade

Yuan

Wang

2019

Remote Sensing

View full text Add to dashboard Cite

A hyperspectral image usually covers a large scale of ground scene, which contains various materials with different spectral properties. When directly exploring the background information using all the image pixels, complex spectral interactions and inter-/intra-difference of different samples will significantly reduce the accuracy of background evaluation and further affect the detection performance. To address this problem, this paper proposes a novel hyperspectral anomaly detection method based on separability-aware sample cascade model. Through identifying separability of hyperspectral pixels, background samples are sifted out layer-by-layer according to their separable degrees from anomalies, which can ensure the accuracy and distinctiveness of background representation. First, as spatial structure is beneficial for recognizing target, a new spectral–spatial feature extraction technique is used in this work based on the PCA technique and edge-preserving filtering. Second, depending on different separability computed by sparse representation, samples are separated into different sets which can effectively and completely reflect various characteristics of background across all the cascade layers. Meanwhile, some potential abnormal targets are removed at each selection step to avoid their effects on subsequent layers. Finally, comprehensively taking different good properties of all the separability-aware layers into consideration, a simple multilayer anomaly detection strategy is adopted to obtain the final detection map. Extensive experimental results on five real-world hyperspectral images demonstrate our method’s superior performance. Compared with seven representative anomaly detection methods, our method improves the average detection accuracy with great advantages.

show abstract

“…It has the spectral range of 370 nm to 2510 nm. After removing some noisy bands [53], totally 207 spectral channels are remained in the experiment. This image has the size of 100 × 100.…”

Section: Data Setsmentioning

confidence: 99%

Section: Data Setsmentioning

confidence: 99%

Hyperspectral Anomaly Detection Based on Separability-Aware Sample Cascade

Yuan

Wang

2019

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…This HS image consists of 126 bands of size 400 × 400 pixels with a spatial resolution of 3.5 m per pixel after removing the noisy bands. The ground truth image has eight classes inside [39]. (4) The last data set is provided by the 2013 Institute of Electrical and Electronics Engineers (IEEE)…”

Section: Study Data Setsmentioning

confidence: 99%

Hyperspectral Image Classification Based on Semi-Supervised Rotation Forest

Lu¹,

Zhang²,

Li³

et al. 2017

Remote Sensing

View full text Add to dashboard Cite

Ensemble learning is widely used to combine varieties of weak learners in order to generate a relatively stronger learner by reducing either the bias or the variance of the individual learners. Rotation forest (RoF), combining feature extraction and classifier ensembles, has been successfully applied to hyperspectral (HS) image classification by promoting the diversity of base classifiers since last decade. Generally, RoF uses principal component analysis (PCA) as the rotation tool, which is commonly acknowledged as an unsupervised feature extraction method, and does not consider the discriminative information about classes. Sometimes, however, it turns out to be sub-optimal for classification tasks. Therefore, in this paper, we propose an improved RoF algorithm, in which semi-supervised local discriminant analysis is used as the feature rotation tool. The proposed algorithm, named semi-supervised rotation forest (SSRoF), aims to take advantage of both the discriminative information and local structural information provided by the limited labeled and massive unlabeled samples, thus providing better class separability for subsequent classifications. In order to promote the diversity of features, we also adjust the semi-supervised local discriminant analysis into a weighted form, which can balance the contributions of labeled and unlabeled samples. Experiments on several hyperspectral images demonstrate the effectiveness of our proposed algorithm compared with several state-of-the-art ensemble learning approaches.

show abstract

“…Özellikle bu görüntülerin sunmuş olduğu yüksek spektral çözünürlük, geleneksel görüntülerin içermediği detaylı spektral özellikleri ortaya çıkarabilmektedir. Bunun sonucunda son yıllarda hedef tespiti [1], hassas tarım uygulamaları [2] ve yer altı maden tespiti [3] gibi çalışmalarda hiperspektral görüntüler akıllı yöntemler ile analiz edilip sınıflandırılarak kullanılmaktadır.…”

Section: Gi̇ri̇ş (Introduction)unclassified

Evrişimsel sinir ağı ve iki-boyutlu karmaşık gabor dönüşümü kullanılarak hiperspektral görüntü sınıflandırma

Hanbay

2019

Gazi Üniversitesi Mühendislik Mimarlık Fakültesi Dergisi

View full text Add to dashboard Cite

Highlights:Graphical/Tabular Abstract  Deep-learned and Gabor filtering-based feature fusion approach is proposed.  A seven layer deep convolutional network is designed to learn deep features.  Classification performance of the proposed hybrid method is comprehensively evaluated.In this paper, a new hyperspectral image classification method based on 2-dimensional complex Gabor filtering and deep convolutional neural network is proposed. To extract a set of features at multiple orientations and frequencies, the 2-D complex Gabor filtering is performed on the first three principal components of the hyperspectral image. The designed deep neural network contains seven layers with weights, including the input layer, the three convolutional layers C1, C3 and C5, the three max pooling layer S2, S4 and S6, the full connection layer F7. In this paper, the deep-learned and Gabor features are simultaneously calculated for the input samples. The proposed Gabor-filtering-assisted deep neural network classifier has been tested on two hyperspectral datasets. The experiments have showed that the proposed deep neural network yields highly accurate classification results in most cases. Figure A. The overall flowchart of the proposed Gabor filtering and ESA-based hyperspectral image classification methodPurpose: The aim of this study is to investigate the contribution of the 2-dimensional complex Gabor transform to deep convolutional neural network architectures. The performance of the proposed hybrid method has been tested on the problem of classification of hyperspectral images. Theory and Methods:Two different kinds of features are calculated. The first one is the deep-learned features. The second one is the Gabor transform features, which are extracted by applying at multiple orientations and frequencies on three principal components of the hyperspectral image. A softmax classifier is trained on the obtained hybrid features for hyperspectral image classification. Results:Hyperspectral images have successfully classified. The proposed Gabor-filtering-assisted deep learning network generally provides higher accuracy than conventional deep learning-based classifier, support vector machine classifier and elastic net representation-based classifier. Conclusion:The experimental results demonstrated that conventional deep learning architectures can still be improved using hybrid feature extraction strategies. The superiority of the proposed method comes from the combination of Gabor transform-based textural features and deep-learned features.

show abstract

Hyperspectral Anomaly Detection With Attribute and Edge-Preserving Filters

Cited by 358 publications

References 37 publications

Hyperspectral Anomaly Detection Based on Separability-Aware Sample Cascade

Hyperspectral Anomaly Detection Based on Separability-Aware Sample Cascade

Hyperspectral Image Classification Based on Semi-Supervised Rotation Forest

Evrişimsel sinir ağı ve iki-boyutlu karmaşık gabor dönüşümü kullanılarak hiperspektral görüntü sınıflandırma

Contact Info

Product

Resources

About