Ruiqing Niu

Anisotropic Diffusion for Hyperspectral Imagery Enhancement

Among all enhancement techniques being developed over the past two decades, anisotropic diffusion has received much attention and experienced significant developments, with promising results and applications in various specific domains. The elegant property of the technique is that it can enhance images by reducing undesirable intensity variability within the objects in the image, while improving SNR and enhancing the contrast of the edges in scalar and, more recently, vector-valued images, such as color, multispectral, and hyperspectral imagery. In this paper, we present an alternative hyperspectral anisotropic diffusion scheme that takes into account the recent advances and the specificities of hyperspectral remote sensing. In addition, the proposed anisotropic diffusion algorithm can improve the classification accuracy of hyperspectral imagery by reducing the spatial and spectral variability of the image, while preserving the edges of objects. It is also revealed that the additive operator splitting scheme of our method can increase computer efficiency. Qualitative experiments, based on a real hyperspectral remote sensing image, show significant improvements in visual effects when using our method. Quantitative analyses, based on classification accuracies, confirm the superiority and validity of the proposed diffusion algorithm.

Region-based adaptive anisotropic diffusion for image enhancement and denoising

A novel region-based adaptive anisotropic diffusion (RAAD) is presented for image enhancement and denoising. The main idea of this algorithm is to perform the region-based adaptive segmentation. To this end, we use the eigenvalue difference of the structure tensor of each pixel to classify an image into homogeneous detail, and edge regions. Ac- cording to the different types of regions, a variable weight is incorporated into the anisotropic diffusion partial differential equation for compromising the forward and backward diffusion, so that our algorithm can adaptively encourage strong smoothing in homogeneous regions and suitable sharp- ening in detail and edge regions. Furthermore, we present an adaptive gradient threshold selection strategy. We suggest that the optimal gradient threshold should be estimated as the mean of local intensity differences on the homogeneous regions. In addition, we modify the anisotropic diffu- sion discrete scheme by taking into account edge orientations. We believe our algorithm to be a novel mechanism for image enhancement and de- noising. Qualitative experiments, based on various general digital images and several T1- and T2-weighted magnetic resonance simulated images, show significant improvements when the RAAD algorithm is used versus the existing anisotropic diffusion and the previous forward and backward diffusion algorithms for enhancing edge features and improving image contrast. Quantitative analyses, based on peak signal-to-noise ratio, the universal image quality index, and the structural similarity confirm the su- periority of the proposed algorithm. C 2010 Society of Photo-Optical Instrumentation

Super-resolution land-cover mapping based on the selective endmember spectral mixture model in hyperspectral imagery

Hyperspectral imagery contains a large number of mixed pixels, which limits its utility. Super-resolution mapping is a potential solution to this problem, designed to use the proportion of land covers to obtain a sharpened thematic map with higher resolution. Endmember is a fundamental variable in the process, which is a critical issue for decomposing the mixed pixels and sharpening the subpixel level images. In most cases, the forms of the endmember combination in diverse pixels are very distinct. However, traditional soft classification methods neglect this point and model endmembers as fixed composition entities. Due to the reliance on this flawed spectral mixture model, the super-resolution mapping is unable to represent detail in the following result image precisely and effectively. In this work, therefore, endmember variability is considered, focusing on identifying the most suitable form of the endmember combination. This issue is addressed by applying a new selective endmember spectral mixture (SESM) model, which allows the endmember number and type to vary at a per pixel level, and then super-resolution mapping can be subsequently performed according to the produced spectral abundances. Two different types of hyperspectral data are used in our experiments. First, the SESM model is tested individually for validation of its applicability. Then the complete algorithm integrating SESM and super-resolution mapping based on a back-propagation neural network is evaluated. It showed that a more accurate endmember combination in the parent pixel results in a finer representation image. The experimental results prove that the proposed algorithm can effectively improve the accuracy of the super-resolution mapping results compared to the traditional method.

The assessment of landslide susceptibility mapping using random forest and decision tree methods in the Three Gorges Reservoir area, China

Landslide susceptibility mapping is an indispensable prerequisite for landslide prevention and reduction. At present, research into landslide susceptibility mapping has begun to combine machine learning with remote sensing and geographic information system (GIS) techniques. The random forest model is a new integrated classification method, but its application to landslide susceptibility mapping remains limited. Landslides represent a serious threat to the lives and property of people living in the Zigui–Badong area in the Three Gorges region of China, as well as to the operation of the Three Gorges Reservoir. However, the geological structure of this region is complex, involving steep mountains and deep valleys. The purpose of the current study is to produce a landslide susceptibility map of the Zigui–Badong area using a random forest model, multisource data, GIS, and remote sensing data. In total, 300 pre-existing landslide locations were obtained from a landslide inventory map. These landslides were identified using visual interpretation of high-resolution remote sensing images, topographic and geologic data, and extensive field surveys. The occurrence of landslides is closely related to a series of environmental parameters. Topographic, geologic, Landsat-8 image, raining data, and seismic data were used as the primary data sources to extract the geo-environmental factors influencing landslides. Thirty-four layers of causative factors were prepared as predictor variables, which can mainly be categorized as topographic, geological, hydrological, land cover, and environmental trigger parameters. The random forest method is an ensemble classification technique that extends diversity among the classification trees by resampling the data with replacement and randomly changing the predictive variable sets during the different tree induction processes. A random forest model was adopted to calculate the quantitative relationships between the landslide-conditioning factors and the landslide inventory map and then generate a landslide susceptibility map. The analytical results were compared with known landslide locations in terms of area under the receiver operating characteristic curve. The random forest model has an area ratio of 86.10%. In contrast to the random forest (whole factors, WF), random forest (12 major factors, 12F), decision tree (WF), decision tree (12F), the final result shows that random forest (12F) has a higher prediction accuracy. Meanwhile, the random forest models have higher prediction accuracy than the decision tree model. Subsequently, the landslide susceptibility map was classified into five classes (very low, low, moderate, high, and very high). The results demonstrate that the random forest model achieved a reasonable accuracy in landslide susceptibility mapping. The landslide hazard zone information will be useful for general development planning and landslide risk management.

Landslide susceptibility mapping based on global and local logistic regression models in Three Gorges Reservoir area, China

This paper investigates the spatial stationarity of the relationship between landslide susceptibility and associated factors in Three Gorges Reservoir area, a landslide-rich area in China. Two logistic regression (LR) models have been used: A global LR (LR) assumes that the regression coefficients remain constant over the whole region, whereas a geographically weighted LR (GWLR) allows the regression coefficients to differ at the local scale. In LR model, lithology seems to have positive influence on the location of landslides, as it has a positive regression coefficient (0.005), while the other factors all have negative effects on landslide susceptibility as they all have negative coefficients. However in GWLR model, lithology does not always keep positive influence, as its coefficients range from −0.533 to 0.695. These results indicate a degree of spatial variation in the relationship between landslide susceptibility and the influencing factors in the study area. Furthermore, six evaluation criteria, based on the fit and complexity of the models, were used to compare the two approaches: deviance, corrected Akaike’s information criterion (AICc), local percent deviance explained (pdev), receiver operating characteristic curve (ROC), Bayesian information criterion (BIC), and residual Moran’s I. The results suggest that GWLR model provides potential advantages in landslide susceptibility mapping and sheds new light on the spatial non-stationarity of the relationship between landslide susceptibility and its influencing factors.

Impact of fractional vegetation cover change on soil erosion in Miyun reservoir basin, China

Abstract It is well known that soil erosion at the watershed scale is the result of interactions between various factors. Among these environmental factors, vegetation is the most important and plays a major role in the soil erosion process. The impact of fractional vegetation cover change (FVCC) on soil erosion in non-contributing areas is a heavily discussed topic. In this paper, the fractional vegetation cover (FVC) in 2002 and 2005 was calculated by using a backpropagation neural network based on remote sensing (RS) data. Then the impacts of FVCC on sediment loads at the outlets of two Miyun reservoir sub-basins were evaluated by integrating RS and geographic information system with statistical analysis. The Miyun reservoir basin (MRB) is characterized by hilly and mountainous topographies and seasonal rainy weather. The primary goal of this paper is to gain a better understanding of FVCC, its driving forces, and its impact on regional soil erosion. We discuss spatiotemporal variations in precipitation and soil erosion, identify which factors contribute to those variations, analyze the influences of FVCC on climate change and human activities and, finally, conclude that changes in FVC and climate regimes are primary factors for soil erosion in MRB. We also discuss how sediment loads may be used to quantitatively separate biophysical and anthropogenic influences and to identify critical thresholds that might have dramatic consequences for the watershed ecosystem. These findings should be quite helpful for sensible watershed development and management planning.

Image restoration and enhancement based on tunable forward-and-backward diffusion

In order to improve signal-to-noise ratio (SNR) and contrast-to-noise ratio, we introduces a novel tunable forward-and-backward (TFAB) diffusion approach for image restoration and edge enhancement. In the TFAB algorithm, an alternative forward-and-backward (FAB) diffusion process is presented, where it is possible to better modulate all aspects of the diffusion behavior and it shows better algorithm behavior compared to the existing FAB diffusion approaches. In addition, there is no necessity to laboriously determine the value of the gradient threshold. We believe the TFAB diffusion to be an adaptive mechanism for image restoration and enhancement. Qualitative experiments, based on various general digital images and a magnetic resonance image, show significant improvements when the TFAB diffusion algorithm is used versus the existing anisotropic diffusion and the previous FAB diffusion algorithms for enhancing edge features and improving image contrast. Quantitative analyses, based on peak SNR and the universal image quality index, confirm the superiority of the proposed algorithm.

A scale-based forward-and-backward diffusion process for adaptive image enhancement and denoising

This work presents a scale-based forward-and-backward diffusion (SFABD) scheme. The main idea of this scheme is to perform local adaptive diffusion using local scale information. To this end, we propose a diffusivity function based on the Minimum Reliable Scale (MRS) of Elder and Zucker (IEEE Trans. Pattern Anal. Mach. Intell. 20(7), 699-716, 1998) to detect the details of local structures. The magnitude of the diffusion coefficient at each pixel is determined by taking into account the local property of the image through the scales. A scale-based variable weight is incorporated into the diffusivity function for balancing the forward and backward diffusion. Furthermore, as numerical scheme, we propose a modification of the Perona-Malik scheme (IEEE Trans. Pattern Anal. Mach. Intell. 12(7), 629-639, 1990) by incorporating edge orientations. The article describes the main principles of our method and illustrates image enhancement results on a set of standard images as well as simulated medical images, together with qualitative and quantitative comparisons with a variety of anisotropic diffusion schemes.

Adaptive pixel unmixing based on a fuzzy ARTMAP neural network with selective endmembers

Pixel unmixing is essential for the reliable description of many land-cover patterns with low spatial resolution. The fuzzy ARTMAP neural network-based model has been proven effective for pixel unmixing in the literature. In most cases, the forms of the endmember combination in diverse pixels are very distinct. However, traditional fuzzy ARTMAP model neglects such difference and models endmembers as fixed composition entities. Due to this limitation, the mixture model is unable to precisely and effectively represent details in the result image. In this work, we address this issue by applying a new selective endmember spectral mixture model based on fuzzy ARTMAP neural network. We first consider the endmember variability and identify the most suitable form of the endmember combination, and then the fuzzy ARTMAP model is used to perform the unmixing work. Through two experiments, we show that a more accurate endmember combination in the parent pixel results in an adaptive representation image. The results confirm that the proposed algorithm can effectively improve the accuracy of the unmixing results compared with the linear unmixing method and the traditional fuzzy ARTMAP model.

Landslide mechanism analysis in the Three Gorges based on cloud model and formal concept analysis

Abstract In the Three Gorges major landslides are the primary disasters, and endanger the normal running of the Three Gorges Dam and the life and property of the residents in the region. Hence, it is very important to formulate effective strategies for the prevention and remediation of landslides in this region, as part of which landslide mechanism analysis is an important task. In this paper, landslide mechanism analysis in the Three Gorges is carried out based on spatial data mining and knowledge discovery. The 1:50000 geological map, 1:10000 relief map and China–Brazil Earth Resources Satellite (Cbers) images were adopted to produce the key factors influencing landslide development, including engineering rock group, reservoir water fluctuation, vegetation coverage, slope structure, elevation, slope and aspect. A soft partition method was adopted to elevate the knowledge levels and formulate the quantification factors qualitatively based on the cloud model. In terms of these qualitative factors, a concept grid is built based on formal concept analysis and a concept grid algorithm. Based on this concept grid, the knowledge related to landslide mechanism is mined from the multi-theme landslide data, including the associations between the various factors that influence a landslide, the circumstances in which a landslide is easily triggered, and the relationship between landslide probability and factor combination. The experimental results show that the knowledge of landslide causes mined by our method possesses high confidence and is in agreement with the field circumstances. Therefore, the spatial data mining method proposed in this paper is suitable for landslide mechanism analysis. It can achieve the transformation between quantitative detection data of landslides and qualitative human mind, thereby leading to an innovative approach for landslide mechanism analysis.