Search

Background subtraction (BS) in video sequences is a main research field, and the aim is to separate moving objects in the foreground from stationary background. Using the framework of schemes-based robust principal component analysis (RPCA), we propose a novel BS method employing the more refined prior representations for the static and dynamic components of the video sequences. Specifically, the rank-1 constraint is exploited to describe the strong low-rank property of background layer (temporal correlation of static component), and 3-D total variation measure and $L_{1}$ norm are used to model the spatial-temporal smoothness of foreground layer and sparseness of noise (dynamic component). This method introduces rank-1, smooth, and sparse properties into the RPCA framework for BS task, and it is dubbed TR1-RPCA. In addition, an efficient algorithm based on the alternating direction method of multipliers is designed to solve the proposed BS model. Extensive experiments on simulated and real videos demonstrate the superiority of the proposed method.

In real hyperspectral scenes, heterogeneous spatial details and noises make a single subspace assumptions unrealistic. In this letter, a novel transferable multiple tensor subspace learning scheme is proposed for super-resolution enhancement of hyperspectral image (HSI). The intrinsic assumption is that the nonlocal patch tensors extracted from HSIs are derived from multiple tensor low-rank subspaces, which is compatible with practical data distribution and may better characterize the complex structures underlying HSIs. The transferable subspace structures are embedded into both nonblind and semi-blind HSI super-resolution. The alternating direction method of multipliers (ADMMs) algorithm is derived for model learning. The superiority of our method is demonstrated by comprehensive experiments on both synthetic and real datasets. [ABSTRACT FROM AUTHOR]

Hyperspectral images (HSIs) often contain pixels with mixed spectra, which makes it difficult to accurately separate the background signal from the anomaly target signal. To mitigate this problem, we present a method that applies spectral unmixing and structure sparse representation to accurately extract the pure background features and to establish a structured sparse representation model at a sub-pixel level by using the Archetypal Analysis (AA) scheme. Specifically, spectral unmixing with AA is used to unmix the spectral data to obtain representative background endmember signatures. Moreover the unmixing reconstruction error is utilized for the identification of the target. Structured sparse representation is also adopted for anomaly target detection by using the background endmember features from AA unmixing. Moreover, both the AA unmixing reconstruction error and the structured sparse representation reconstruction error are integrated together to enhance the anomaly target detection performance. The proposed method exploits background features at a sub-pixel level to improve the accuracy of anomaly target detection. Comparative experiments and analysis on public hyperspectral datasets show that the proposed algorithm potentially surpasses all the counterpart methods in anomaly target detection.

Loss of lean body mass is a relevant component of the cachexia, or protein energy wasting (PEW), syndrome. Reduced muscle mass seems to be the most solid criterion for the presence of cachexia/PEW in patients with chronic kidney disease (CKD), and those with greater muscle mass loss have a higher risk of death. Children with CKD have many risk factors for lean mass and muscle wasting, including poor appetite, inflammation, growth hormone resistance, and metabolic acidosis. Mortality risks in patients with CKD increases as body mass index (BMI) and weight decreases. However, data regarding cachexia/PEW and muscle wasting in children with CKD is scarce due to lack of consensus in diagnostic criteria and an appropriate investigative methodology. Further research is urgently needed to address this important complication in the pediatric CKD setting, which may have fundamental impact on clinical outcomes., Author(s): Eduardo A. Oliveira [sup.1] [sup.2] , Wai W. Cheung [sup.1] , Kalodiah G. Toma [sup.1] , Robert H. Mak [sup.1] Author Affiliations: (Aff1) 0000 0001 2107 4242, grid.266100.3, Division [...]

The present research developed a novel methodological framework to differentiate natural mangrove stands (i.e., original), from stands which were planted and stands naturally established after interaction between planted and non-planted stands (e.g., through pollination, i.e., non-original). Ground-truth and remote sensing data were collected for Zhanjiang Mangrove National Nature Reserve (ZMNNR) in P.R. China. First, satellite images of Corona (1967) and GeoEye-1 (2009) were overlaid to identify original (1967) and non-original (2009) mangrove stands. Second, in both stands a total of 75 in situ plots (25 m2) were measured for ground-truthing of tree structural parameters including height, density, basal area and Complexity Index (CI). From temporal satellite data, we identify 236.12 ha of original mangrove and 567.88 ha of non-original mangrove in the reserve. Averaged measurements of the original mangrove stands, i.e., stem density (1164 nos. 0.1 ha−1), basal area (90.3 m2 0.1 ha−1) and CI (100.59), indicated that they were in a state of maturity and less disturbed compared to the non-original mangroves (density, 1241 nos. 0.1 ha−1; basal area, 4.92 m2 0.1 ha−1 and CI, 55.65). The Kruskal–Wallis test showed significant differentiation between the original and non-original mangrove tree structural parameters: Kandelia obovata’s density, X2 = 34.78, d.f. = 1, p = 0.001; basal area, X2 = 108.15, d.f. = 1, p = 0.001; Rizhopora stylosa’s density, X2 = 64.03, d.f. = 1, p = 0.001; basal area, X2 = 117.96, d.f. = 1, p = 0.001. The latter is also evident from the clustering plots generated from the Principal Component Analysis (PCA). Vegetation dynamics at the ZMNNR also enabled us to compare the species composition and distribution patterns with other Indo-West Pacific regions. Overall, the present study not only highlights the advantage of >50 years old satellite data but also provide a benchmark for future ecological research, conservation and management of the ZMNNR.

The Instant Dehaze method used polarized images to obtain a dehazed image and an estimated depth map of the scene. Haze due to atmospheric absorption and scattering causes degradation in image quality and the estimated depth. This estimated depth is misrepresented due to high degree of polarization and scene's objects directly illuminated by the sun. In this paper, a polarization guided autoregressive (AR) model for depth recovery is presented. This proposed method restores the estimated depth map by incorporating polarized data to an adaptive AR model. First a 90° polarized image is used in our polarization term of AR coefficient, then the Stokes vector component S1 is used in our polarization guided depth map in the depth term of AR coefficient. The experimental results show that our method outperforms existing state-of-the-art schemes and improves conventional polarization dehazing method.

This paper presents a correlation filter object tracker based on fast spatial-spectral features (FSSF) to realize robust, real-time object tracking in hyperspectral surveillance video. Traditional object tracking in surveillance video based only on appearance information often fails in the presence of background clutter, low resolution, and appearance changes. Hyperspectral imaging uses unique spectral properties as well as spatial information to improve tracking accuracy in such challenging environments. However, the high-dimensionality of hyperspectral images causes high computational costs and difficulties for discriminative feature extraction. In FSSF, the real-time spatial-spectral convolution (RSSC) kernel is updated in real time in the Fourier transform domain without offline training to quickly extract discriminative spatial-spectral features. The spatial-spectral features are integrated into correlation filters to complete the hyperspectral tracking. To validate the proposed scheme, we collected a hyperspectral surveillance video (HSSV) dataset consisting of 70 sequences in 25 bands. Extensive experiments confirm the advantages and the efficiency of the proposed FSSF for object tracking in hyperspectral video tracking in challenging conditions of background clutter, low resolution, and appearance changes.

Conventional classification algorithms have shown great success in balanced hyperspectral data classification. However, the imbalanced class distribution is a fundamental problem of hyperspectral data, and it is regarded as one of the great challenges in classification tasks. To solve this problem, a non-ANN based deep learning, namely SMOTE-Based Weighted Deep Rotation Forest (SMOTE-WDRoF) is proposed in this paper. First, the neighboring pixels of instances are introduced as the spatial information and balanced datasets are created by using the SMOTE algorithm. Second, these datasets are fed into the WDRoF model that consists of the rotation forest and the multi-level cascaded random forests. Specifically, the rotation forest is used to generate rotation feature vectors, which are input into the subsequent cascade forest. Furthermore, the output probability of each level and the original data are stacked as the dataset of the next level. And the sample weights are automatically adjusted according to the dynamic weight function constructed by the classification results of each level. Compared with the traditional deep learning approaches, the proposed method consumes much less training time. The experimental results on four public hyperspectral data demonstrate that the proposed method can get better performance than support vector machine, random forest, rotation forest, SMOTE combined rotation forest, convolutional neural network, and rotation-based deep forest in multiclass imbalance learning.

The accuracy of anomaly detection in hyperspectral images (HSIs) faces great challenges due to the high dimensionality, redundancy of data, and correlation of spectral bands. In this paper, to further improve the detection accuracy, we propose a novel anomaly detection method based on texture feature extraction and a graph dictionary-based low rank decomposition (LRD). First, instead of using traditional clustering methods for the dictionary, the proposed method employs the graph theory and designs a graph Laplacian matrix-based dictionary for LRD. The robust information of the background matrix in the LRD model is retained, and both the low rank matrix and the sparse matrix are well separated while preserving the correlation of background pixels. To further improve the detection performance, we explore and extract texture features from HSIs and integrate with the low-rank model to obtain the sparse components by decomposition. The detection results from feature maps are generated in order to suppress background components similar to anomalies in the sparse matrix and increase the strength of real anomalies. Experiments were run on one synthetic dataset and three real datasets to evaluate the performance. The results show that the performance of the proposed method yields competitive results in terms of average area under the curve (AUC) for receiver operating characteristic (ROC), i.e., 0.9845, 0.9962, 0.9699, and 0.9900 for different datasets, respectively. Compared with seven other state-of-the-art algorithms, our method yielded the highest average AUC for ROC in all datasets.