Remote sensing technologies, such as aerial photography and satellite remote sensing, play a crucial role in assessing and monitoring the changes in land use and land cover (LULC) for large areas. Deep learning methods are recently becoming popular for LULC classification and have maintained promising performance in many applications. However, challenges remain in LULC classification for the complex semantic appearance in high-resolution remote sensing images. In this paper, we proposed a novel framework that cascades two deep learning models i.e., ResNet and SegNet for land use and land cover classification. The advantages of strong semantic feature extraction (ResNet) and efficient boundary delineation (SegNet) capabilities of these two models can be combined to derive highly accurate land use and land cover classification results. Our method performed well on the recently developed multisource Dense-Pixel Annotation Dataset Globe230K.

The expansion of marine transportation, along with the growth of offshore oil exploration and extraction, has elevated the risk of marine oil spill incidents which represents a serious threat to the marine environment. Polarimetric SAR (PolSAR) has proven its advantage in distinguishing clean seawater, mineral oil spills, and its biogenic look-alikes. However, the number of quad polarimetric Multi-Look Complex (MLC) UAVSAR images with verified marine oil spills is very limited. Meanwhile, plenty of SAR oil spill sample images are in the single polarimetric format. Therefore, it is a promising approach to jointly use the semantic information provided by single-pol SAR dataset and the polarimetric backscattering information provided by the quad-pol dataset to improve the accuracy of oil spill detection. In response, this study proposes a novel study of feature concatenation (features fusion) technique to integrate information from two of these different sources. It combines the capabilities of the U-Net model, which processes single-pol SAR images, with those of a 9-channels RV (Real Valued) U-Net model, which handles quad-pol images. The experiment preliminarily proves the effectiveness of the proposed method.

Marine oil spills have caused serious harm to the costal ecological environment and marine economy. Synthetic aperture radar (SAR) has become a major equipment for oil spill detection because of its advantages of all-day and all-weather observation capability. In this paper, the complex convolutional neural network (CVCNN) framework is applied for marine oil spills classification. The classification performance of different polarization modes on marine oil spills classification is analyzed. Experimental results show that CP SAR modes have comparable performance as QP mode for marine oil spills classification. Among them, the Circular Transmit and Linear Receive (CTLR) mode has the best classification performance in the framework of CVCNN while DP (VVVH) mode is another promising alternative with much simpler hardware configuration requirement.

The brightness temperature (TB) features extracted from Chang’e Lunar Microwave Sounder (CELMS) data have been proved their superiority to study mare basalt. In this paper, dimension reduction and feature space analysis are conducted on TBfeatures to fully understand the data distribution and reduce the feature redundancy in the classification process based on two methods - Principal Component Analysis (PCA) and Nonnegative Matrix Factorization (NNMF). The results showed that PCA and NNMF can effectively enhance the classification capability of early(?)- and late(?)-age mare basalt respectively, and proved the necessity of dimension reduction for CELMS TBfeatures due to the largely-existing redundancy.

Abstract Automatic segmentation of optic disc (OD) and optic cup (OC) is an essential task for analysing colour fundus images. In clinical practice, accurate OD and OC segmentation assist ophthalmologists in diagnosing glaucoma. In this paper, we propose a unified convolutional neural network, named ResFPN-Net, which learns the boundary feature and the inner relation between OD and OC for automatic segmentation. The proposed ResFPN-Net is mainly composed of multi-scale feature extractor, multi-scale segmentation transition and attention pyramid architecture. The multi-scale feature extractor achieved the feature encoding of fundus images and captured the boundary representations. The multi-scale segmentation transition is employed to retain the features of different scales. Moreover, an attention pyramid architecture is proposed to learn rich representations and the mutual connection in the OD and OC. To verify the effectiveness of the proposed method, we conducted extensive experiments on two public datasets. On the Drishti-GS database, we achieved a Dice coefficient of 97.59%, 89.87%, the accuracy of 99.21%, 98.77%, and the Averaged Hausdorff distance of 0.099, 0.882 on the OD and OC segmentation, respectively. We achieved a Dice coefficient of 96.41%, 83.91%, the accuracy of 99.30%, 99.24%, and the Averaged Hausdorff distance of 0.166, 1.210 on the RIM-ONE database for OD and OC segmentation, respectively. Comprehensive results show that the proposed method outperforms other competitive OD and OC segmentation methods and appears more adaptable in cross-dataset scenarios. The introduced multi-scale loss function achieved significantly lower training loss and higher accuracy compared with other loss functions. Furthermore, the proposed method is further validated in OC to OD ratio calculation task and achieved the best MAE of 0.0499 and 0.0630 on the Drishti-GS and RIM-ONE datasets, respectively. Finally, we evaluated the effectiveness of the glaucoma screening on Drishti-GS and RIM-ONE datasets, achieving the AUC of 0.8947 and 0.7964. These results proved that the proposed ResFPN-Net is effective in analysing fundus images for glaucoma screening and can be applied in other relative biomedical image segmentation applications.

The recent development of marine transportation and offshore oil exploration and exploitation has increased the risk of marine oil spill accidents. Oil pollution is one of the most complex marine pollutions, it will seriously threaten the marine ecological environment. Synthetic aperture radar (SAR) has been widely used in marine monitoring for its all-day and all-weather imaging capability. Polarimetric SAR can obtain polarimetric scattering information of the ground targets, which provides a new means for marine oil spill detection. Besides, with the recent development of machine learning algorithms, especially convolutional neural networks, higher oil spill classification accuracy can be obtained given more training datasets. However, currently most neural network models are applied to real-valued input and cannot fully exploit the phase information contained in complex-valued data of polarimetric SAR images. In this study, a classification approach of marine oil spills in polarimetric SAR images is presented based on the complex-valued convolutional neural network (CVCNN). The experimental results show that the proposed approach outperforms the real-valued convolutional neural network (RVCNN) in both the detection of oil spills from sea surface and the classification of crude oil films and biogenic oil films.

Abstract Histogram equalization (HE) is a classic and widely used image contrast enhancement algorithm for its good performance and high efficiency. However, over‐enhancement caused by high peak in the histogram affects the subjective quality of the image processed by HE to a large extent. In this paper, a detail weighted histogram equalization (DWHE) method is proposed based on a novel histogram modification (HM) model named partial statistic and global mapping (PSGM) to alleviate high peak and suppress over‐enhancement. Moreover, the authors implement a refined version of gamma correction (GC) named texture enhancement function (TEF) on high‐frequency images to reduce the noise amplification effect. At last, the authors propose a novel adaptively weighted pixel‐level image fusion method to further reduce the phenomenon of over‐enhancement and improve brightness distribution. Both subjective and quantitative evaluations are conducted on images containing a variety of scenes. Compared with several state‐of‐the‐art image enhancement methods, the proposed framework obtained generally the best performance in aspects of both subjective appearance and objective evaluation indices. Therefore, it is proved that the proposed methods can effectively alleviate the over‐enhancement, enrich image details, and efficiently improve the visual quality while preserving the brightness of the image.

Emotion recognition from macroexpression and microexpression has been widely used in applications such as human–computer interaction, learning status evaluation, and mental disorder diagnosis. However, due to the complexity of human macroexpressions, recognizing macroexpressions with high accuracy is a challenging task. Moreover, the short duration and low movement intensity of microexpressions make its recognition more difficult. For MM-FER (macro and microfacial expression recognition), the key information can be more efficiently expressed by a graph. In this article, a novel framework based on graph neural network named SSGNN (spatial and spectral domain features based on a graph neural network) is designed to extract spatial and spectral domain features from facial images for MM-FER, which can efficiently recognize both macroexpressions and microexpressions under the same model. SSGNN consists of two parts, SPAGNN and SPEGNN, which are used to extract spectral and spatial domain features, respectively. Experiments proved that jointly using the spectral and spatial information extracted by SSGNN can largely improve the performance of MM-FER when the training sample is limited. First, the influences of different neighbors and samples to the model performance was analyzed. Then, the contribution of SPAGNN and SPEGNN were evaluated. It was discovered that fusing the result of SPAGNN and SPEGNN at decision level further improved the performance of MM-FER. Experiment proved that SSGNN can recognize microexpression acquired by various sensors with higher accuracy under different image resolutions and image formats than the compared state-of-the-art methods in most cases. A cross-dataset experiment demonstrated the generalization ability of SSGNN.

Urban impervious surfaces (UIS) indicate the environmental and socioeconomic influences of rapid urbanization. Synthetic aperture radar (SAR) reflects the scattering behaviors of different land covers while multispectral data demonstrate their physicochemical properties. Numerous studies reported that the incorporation of SAR and optical data supplement each other for better extracting UIS, nevertheless, the shadow and layover effects remain unclear, especially in very high-resolution observations. This study analyzed the shadow and layover influences from both optical and SAR data for fine resolution UIS estimation. Given the SAR shadow and layover distribution, we proposed a multisource shadow-based fuzzy set (MSFS) approach for fusing optical and SAR in optical shadow areas using decision fusion. SAR layovers showed effectiveness in UIS extraction. MSFS delivered 3% and 7% improvement in overall accuracy compared with SVM and RF using feature fusion respectively.

The extraction of urban impervious surface information plays a key role in the studies of urbanization and its related environmental issues. Optical and SAR remote sensing provides complementary information to improve the accuracy of impervious mapping. However, the fusing of information acquired by different sensors is challenging. Optical and SAR features have distinct characteristics, and require different classification strategy and classification types. In this study, a strategy of fusing multi-spectral optical and polarimetric SAR data at decision-level is proposed. Features are extracted from optical and SAR data, then staked auto-encoder is applied to achieve the land use and land cover classification separately. D-S evidence theory is used to fuse the classification result and the imperious surface map is derived. The experiment was conducted in a highly complex urban area of Hong Kong and the results proves the soundness of the method.

In this paper, a unified framework was applied to derive compact polarimetric Synthetic Aperture Radar (SAR) features under general transmit and linear receive polarization conditions. Following the rationale of polarization signature, the characteristics of features derived by transmission of variant roll angles and ellipticity are analyzed. Statistical distance was proposed to quantitatively measure the performance of these compact polarimetric SAR features on marine oil spill classification. Experiment was conducted on a Radarsat-2 quad-pol SAR scene acquired during a controlled oil-on-water exercise.

This study proposed a novel method to analyze slick-free and oil covered sea surface backscattering in the special case of Deepwater Horizon (DWH) oil spill accident based on combination of tilted Bragg scattering and volume scattering components. The DWH accident represents a particular and challenging case due to the very large amount of leaked oil that came from the bottom of the ocean. The proposed scattering model consists of first estimate the large scale tilting angle of sea surface Bragg scattering mechanism and then of the retrieval, through its linear relationship with the relative dielectric constant, of oil-insea volume concentration. Finally, Bragg and volume scattering components can be estimated which provide useful information for a better understanding of i) the sea surface status and ii) the weak-damping properties the leaked oil due to purification/emulsification phenomena. The model is tested considering actual UAVSAR L-band fully-polarimetric SAR data.

Oil spill has been a crucial hazard to the coastal environment. A major difficulty of Synthetic Aperture Radar (SAR) based oil-spill detection algorithms is the classification between mineral oil and biogenic look-alikes. Polarimetric SAR features provides helpful information in disguising mineral oil and its look-alikes. In this study, we focused on the extraction and selection of fully polarimetric SAR features for the classification between mineral and biogenic look-alikes. Three mainly used supervised classifiers including Support vector machine (SVM), Artificial neural network (ANN) and Maximum likelihood classification (ML) were comparatively studied. In the experiment, classification performance increases with the growth of the feature number initially, but still fluctuates or decreases after the sufficient features are considered. It was also discovered that among all the classifiers, support vector machine performed best.

Land use/land cover (LULC) classification using optical and synthetic aperture radar (SAR) remote sensing images is becoming increasingly important to produce more accurate LULC products. As an important step, feature normalization techniques have been studied by the areas of pattern recognition. Nevertheless, because of the totally different imaging mechanisms of optical and SAR sensors, most of the existing normalization approaches are not suitable for optical and SAR data fusion. Moreover, whether normalization is a significant step remains unclear regarding optical and SAR fusion. Taking the Satellite Pour l'Observation de la Terre (SPOT-5) and the Advanced Land Observing Satellite (ALOS)/Phased Array type L-band SAR (PALSAR) (HH and HV polarizations) as the optical and SAR data, this letter aims to evaluate the impact of feature normalization. Experimental results indicated that feature normalization is not necessarily significant depending on fusion methods. For instance, distribution-dependent classifiers (e.g., a maximum likelihood classifier) are independent of feature normalization; thus, it has no impact on the results when using these classifiers. Moreover, advanced classifiers (e.g., a support vector machine) with built-in normalization are also not influenced by feature normalization. In contrast, a minimum distance classifier and an artificial neural network (ANN) depend on the input values of optical and SAR features and thus can be influenced by feature normalization. However, our experiments showed a fluctuation in classification accuracy using an ANN with normalized features. Therefore, more experiments are required to investigate the optimal normalization approaches for the optical and SAR images when using an ANN as the fusion method.

Geladandong Ice Field is one of the largest ice fields in central Qinghai-Tibetan Plateau, also the water source of Yangtze River, Selin Co Lake and Chibozhang Co Lake. In this study, we aim to monitor the glacier flow rates and its change in 1990s and 2000s by Differential SAR Interferometry and offset tracking. We obtained SAR images acquired by ERS-1/2 and Envisat/ASAR in 1990s and 2000s then processed them with offset-tracking method, also validated with D-InSAR method. The result indicates that most glaciers in Geladandong Ice Fields kept stable flow rates at 15-30m/a during 1990s and 2000s. Some glaciers are identified as surge type. We selected two glaciers and studied the time series of flow velocity profiles. During the surging period, flow rate could be 3 to 10 time as normal years. These two glaciers forwarded their terminus during their surging period by Landsat optical monitoring. We concluded that during 1990s and 2000s most glaciers kept a stable flow velocity and some glacier surged during the study period therefore terminus forwarding cannot equal to mass gaining.

Polarimetric features retrieved from CTLR (circularly transmit and linearly receive) Synthetic Aperture Radar (SAR) data was analysed in details. A new parameter, namely, Damping Status Sensitivity Index (DSSI) was proposed for quantitatively evaluating the PolSAR characteristics' capability of discriminating different damping patterns of oil slicks and clean seawater. The L-band Uninhabited Aerial Vehicle SAR (UAVSAR) data was utilized in the experiments. It was shown that polarimetric characteristics retrieved from CTLR compact polarimetric SAR data were nearly as same as those derived from fully polarimetric SAR data and can be applied for discriminating different damping status of oil spill and look-likes.

In this paper a polarimetric SAR data study is undertaken over the critical and peculiar case of the Deepwater Horizon (DWH) oil spill accident occurred in the Gulf of Mexico in 2010. It is a complex case that still calls for deeper physical/chemical characterization due to its special nature. Such an oil spill accident is in fact related to a deep/ultra-deep sea drilling, that although of paramount interest in gas and oil exploitation, need special care and safety actions to manage high oil pressure and potential vast environmental oil spill long-term and short-term impact.

Synthetic Aperture Radar (SAR) data has been reported to be able to provide complementary information towards optical remote sensing data for improving the urban impervious surface estimation. However, most existing researches were focused on using only single polarization SAR data. This study presents a preliminary experiment on the combined use of multispectral optical data and dual polarization SAR data for impervious surfaces estimation. Experimental results using SPOT-5 and ALOS PALSAR images showed a consistent result compared with our previous result using single polarization SAR data. Comparison results showed that not every polarimetric feature was able to provide positive effect to the impervious surfaces estimation. Compared with using only optical and SAR data, the separated HH and HV polarization data provided a positive effect to the result by improving the accuracy. The incorporation of both Entropy and Alpha features were also able to improve the accuracy. However, the HH/HV ratio and the separated use of Entropy did not provide positive results. A combination of all the features turned out to obtain the highest accuracy compared with using a subset of the features.

An improved reconstruction algorithm is proposed for compact polarimetric (CP) synthetic aperture radar (SAR) on oil spill detection. Based on the differences in statistical behavior between open and oil-covered sea surfaces, the proposed algorithm can iteratively reconstruct quad-pol SAR images from CP SAR data. During the experiment, it outperformed two existing compact SAR reconstruction algorithms in terms of both statistical and information theoretical analysis.

The compact polarimetric mode of synthetic aperture radar (SAR) provides an approach of polarimetric observation by a reduced PRF and doubled swath width, which has already proven its advantage compared with the traditional full-polarimetry (FP) mode in several occasions. In this paper, a quantitative evaluation method for the compact polarimetric SAR image reconstruction is proposed. The quad-pol data are reconstructed by the iteration algorithm proposed by Souyris et al.. Analysis is carried out on the performance of the algorithm based on information theoretic analysis. The performance of the algorithm is evaluated quantitatively by implementing information entropy, mutual information, Kullback-Leibler distance and difference of entropy between the reconstructed and original FP data. Based on this analysis, several key parameters of this algorithm is verified and optimized.