Matting model guided variational pansharpening with fractional order gradient transferring

ABSTRACT

In this letter, we proposed a matting model guided variational pansharpening method with fractional order gradient transferring constraint to fuse a low-resolution (LR) multispectral (MS) and a high-resolution (HR) panchromatic (Pan) image to an HR-MS image. Specifically for the proposed model, we not only used the commonly local spectral fidelity constraint for spectral information preserving, but also applied the famous matting model fusion method to further enforce a spatial-spectral fidelity constraint, and particularly proposed a fractional-order gradient transferring constraint to transfer the spatial information of Pan image to the desired HR-MS image. Then, we designed an optimization algorithm to solve the proposed model efficiently. Finally, the experimental results showed the superiority of the proposed method to various methods subjectively and objectively.     

A pansharpening method based on modified cartoon plus texture decomposition

In this paper, a new pansharpening method is proposed to obtain high resolution multi spectral image, while preserving spectral signature. Spectral information is considered to be a piecewise smooth area with almost clear boundaries, and is represented in the cartoon space. Spatial information is described by the orthogonal complement of the cartoon space, which is the texture space. Therefore, remote sensing (RS) images can be transferred to a new space by the cartoon plus texture (CPT) decomposition, where spatial and spectral information can be discriminated properly, and pansharpening is done by substituting multispectral texture components with panchromatic texture component. However, cartoon components of the multispectral bands remains unchanged, and the edges and boundaries of the fused image do not sharpen enough. Therefore, we propose a new decomposition model, using the gradient of the panchromatic cartoon component, to fortify boundaries and some important details, while preserving spectral quality. The proposed method is compared with the well-known classic pansharpening methods. Experimental results show that the proposed method has a better performance in terms of both spatial and spectral qualities, however, it is not efficient in terms of computing time.     

Spectral-spatial constrained sparse unmixing of hyperspectral imagery using a hybrid spectral library

Spectral unmixing based on sparse regression model has recently attracted the hyperspectral data processing community. The technology has two important characteristics. First, sparseness of the abundance matrix is induced because the number of materials in a mixed pixel is very small when compared with the signatures in spectral library. Second, constraints are applied to improve the reconstruction accuracy of the optimization problem. In this letter, a spectral and spatial constrained sparse unmixing (SSCSUn) method is proposed, in which the total variation (TV) regularization with a spatial weight factor is incorporated to exploit spatial information more effectively. Furthermore, several endmembers from the real image are used to form a new hybrid spectral library and are applied as spectral a priori constraints. Experimental results on simulated and real hyperspectral images demonstrate the effectiveness of the proposed approach.     

Spectral–spatial classification of hyperspectral images using deep convolutional neural networks

In this letter, a novel deep learning framework for hyperspectral image classification using both spectral and spatial features is presented. The framework is a hybrid of principal component analysis, deep convolutional neural networks (DCNNs) and logistic regression (LR). The DCNNs for hierarchically extract deep features is introduced into hyperspectral image classification for the first time. The proposed technique consists of two steps. First, feature map generation algorithm is presented to generate the spectral and spatial feature maps. Second, the DCNNs-LR classifier is trained to get useful high-level features and to fine-tune the whole model. Comparative experiments conducted over widely used hyperspectral data indicate that DCNNs-LR classifier built in this proposed deep learning framework provides better classification accuracy than previous hyperspectral classification methods.     

A spectral–spatial similarity-based method and its application to hyperspectral image classification

In this letter, a novel method using both spectral and spatial information is proposed for hyperspectral image classification. Image pixels are partitioned into two sets: a labelled set and an unlabelled set. The goal of this method is to label all the unlabelled pixels. The proposed technique consists of two steps. In the first step, a similarity-based model, in the spectral domain, computes the probability that an unlabelled pixel has the same label as a labelled pixel. In order to improve the classification accuracy, we provide a powerful way to account for spatial information in the second step. Evaluation of the developed method is done on hyperspectral images. Experimental results are compared with those obtained using other hyperspectal image classification methods. The proposed approach performs better than the other ones in terms of classification accuracy.     

Subpixel mapping of urban built-up areas using spatial-spectral information from satellite multispectral remote sensing imagery

ABSTRACT

A novel method for subpixel mapping of urban built-up areas (SMUBA) using spatial-spectral information from satellite multispectral remote sensing imagery is proposed. SMUBA contains two terms, a spatial term and a spectral term. First, a spatial attraction model is used to produce the spatial term. The Normalized difference built-up index is then utilized to obtain the spectral term. Finally, a particle swarm optimization algorithm is used to optimize the two terms and obtain the final mapping result. Since the spatial-spectral information of the urban built-up areas from multispectral imagery is fully utilized by the two terms, the final mapping result is improved. Experimental results on two Landsat 8 Operational Land Imager data show that SMUBA produces better mapping results than state-of-the-art subpixel mapping methods. Moreover, the proposed method does not need any prior shape information.     

A novel variational model for pan-sharpening based on L1 regularization

The Pan-sharpening process synthesizes a multispectral image of high spatial and high spectral quality by injecting the spatial details extracted from the high spatial resolution panchromatic image into the low-resolution multispectral image. In this letter, we propose a novel variational model for pan-sharpening based on L1 regularization, which consists of three energy terms. The first energy term extracts the geometric information from panchromatic image with exponential enhancement function, which can improve the contrast of the fused image by nonlinear expanding of gradient magnitude. The second energy term develops the low-pass filter based on the modulation transfer function (MTF) of the different MS sensor, which can preserve spectral information by determining the injected spatial details adaptively. Compared with the total variation regularization, the L1 regularization encourages the feature of sparse representation, which can improve the spatial fidelity of pan-sharpened images. In addition, introducing L1 regularization term into the variational framework can smooth the image noise and guarantee the stability of the numerical solution. Experimental results demonstrate that the proposed method outperforms several state-of-the-art pan-sharpening methods, including intensity-hue-saturation combined with Brovey transform (IHS-BT), additive wavelet luminance propotional (AWLP), total variation regularization (TVR) and MTF Contrast-based (MTF-CON) method.     

A novel change detection method combined with registration for SAR images

In this letter, we propose a novel change detection method combined with image registration. Inspired by the mutual beneficial property of change detection and image registration, we model these two tasks into a unified process and represent this process as an optimization problem. Then, we use the alternating direction method (ADM) algorithm with proper initialization to solve the optimization problem. Finally, we use spatial information to refine the changed component obtained by the ADM algorithm. Experimental results accomplished over synthetic data and RADARSAT-2 synthetic aperture radar (SAR) images demonstrate that the proposed method can improve both the accuracy of image registration and change detection.     

Advanced Markov random field model based on local uncertainty for unsupervised change detection

Markov random field (MRF)-based methods are effective and popular unsupervised methods for detecting changes in remotely sensed images. In this method, the spatial contextual information is well utilized to conquer the problem of noise sensitivity in the pixel-wise change detection methods. Meanwhile, MRF also suffers from the over-smooth problem and the hard balance between denoising and detail preserving. To tackle these limitations, this letter presented an advanced MRF model based on local uncertainty (LUMRF). First, fuzzy c-means (FCM) cluster method is applied to the difference image obtained by change vector analysis to character each pixel with an initial label (change or no-change) and the corresponding membership values. To improve the detail preservation ability of MRF, the local uncertainty in a given window is subsequently computed and then integrated in the spatial energy term of MRF model. Finally, a refined change map is produced by the proposed LUMRF method. Two experiments were conducted to evaluate the effectiveness of the proposed method. The results show that, in comparison to FCM and MRF, LUMRF gives a better performance with the lowest total error detection and the performance is more robust to the parameter changes.     

Application of fractional-order differentiation in multispectral image fusion

In this letter, a novel pansharpening method is proposed using component substitution (CS) framework. In order to inject the spatial details into the low resolution multispectral (MS) bands, the fractional-order differentiation is used. Eight direction masks are superimposed on each other to construct a unique mask. The primitive detail map is calculated using the difference between the panchromatic (PAN) image and a linear combination of the low resolution MS bands. To refine the detail map and better pansharpening, the superimposed mask is convolved with the extracted primitive detail map. Two datasets collected by the WorldView-2 and Pleiades satellites are used to examine the proposed method. Experimental results show that in comparison with the state-of-the-art methods, the proposed method can better provide the spectral and spatial information in the fused product quantitatively and subjectively.     

A subpixel mapping algorithm combining pixel-level and subpixel-level spatial dependences with binary integer programming

A new subpixel mapping (SPM) algorithm combining pixel-level and subpixel-level spatial dependences is proposed in this letter. The pixel-level dependence is measured by the spatial attraction model (SAM) with either surrounding or quadrant neighbourhood, while the subpixel-level dependence is characterized by either the mean filter or the exponential weighting function. Both pixel-level and subpixel-level dependences are then fused as the weighted dependence in the constructed objective function. The branch-and-bound algorithm is employed to solve the optimization problem, and thus, obtain the optimal spatial distribution of subpixel classes. An artificial image and a set of real remote sensing images were tested for validation of the proposed method. The results demonstrated that the proposed method can achieve results with greater accuracy than two traditional SPM methods and the mixed SAM method. Meanwhile, the proposed method needs less computation time than the mixed SAM, and hence it provides a new solution to subpixel land cover mapping.     

Spectral-Spatial classification of hyperspectral images using functional data analysis

In this letter, a spectral-spatial classification method using functional data analysis (FDA) is proposed. Since the efficacy of the FDA for hyperspectral image analysis instead of analysis in multivariate analysis framework (MAF) was proved previously, we apply FDA to better extract spectral and spatial information for hyperspectral image classification. Therefore, in the FDA framework a support vector machine (SVM) classifier is used for hyperspectral image classification and a watershed segmentation algorithm is applied in order to extract spatial structures. Several approaches to figure a one-band gradient image, as an input to watershed transformation, are examined and investigated. As a result, the extracted segmentation map is used to improve the pixel-wise classification accuracy on which the classification and the segmentation results are combined together using majority vote approach. The efficiency of the proposed method is evaluated on two hyperspectral data sets. The experimental results show that the proposed spectral-spatial classification method provides better classification accuracies compared to some state-of-the-art spectral-spatial classification methods.     

A deep learning framework for hyperspectral image classification using spatial pyramid pooling

In this letter, a new deep learning framework for spectral–spatial classification of hyperspectral images is presented. The proposed framework serves as an engine for merging the spatial and spectral features via suitable deep learning architecture: stacked autoencoders (SAEs) and deep convolutional neural networks (DCNNs) followed by a logistic regression (LR) classifier. In this framework, SAEs is aimed to get useful high-level features for the one-dimensional features which is suitable for the dimension reduction of spectral features, while DCNNs can learn rich features from the training data automatically and has achieved state-of-the-art performance in many image classification databases. Though the DCNNs has shown robustness to distortion, it only extracts features of the same scale, and hence is insufficient to tolerate large-scale variance of object. As a result, spatial pyramid pooling (SPP) is introduced into hyperspectral image classification for the first time by pooling the spatial feature maps of the top convolutional layers into a fixed-length feature. Experimental results with widely used hyperspectral data indicate that classifiers built in this deep learning-based framework provide competitive performance.     

Mutual information based multi-modal remote sensing image registration using adaptive feature weight

Multi-module images registration is a challenging task in image processing, and more especially in the field of remote sensing. In this letter, we strive to present a novel mutual information scheme for image registration in remote sensing scenario based on feature map technique. We firstly take saliency detection advantages to extract geographic pattern, and then utilize the efficient Laplacian of Gaussian(LOG) and Guided Filter methods to construct a new feature map based on different characteristic of multi-channel images. To avoid practical traps of sub-optimization, we propose an novel mutual information(MI) algorithm based on an adapted weight strategy. The proposed model divides an image into patches and assigns weighted values according to patch similarities in order to solve the optimization problem, improve accuracy and enhance performance. Note that, our proposed method incorporates the LOG and Guided Filter methods into image registration for the first time to construct a new feature map based on differences and similarities strategy. Experiments are conducted over island and coastline scenes, and reveal that our hybrid model has a significant performance and outperforms the state-of-the-art methods in remote sensing image registration.     

A novel dynamic threshold method for unsupervised change detection from remotely sensed images

In this letter, a dynamic threshold method is proposed for unsupervised change detection from remotely sensed images. First, change vector analysis technique is applied to generate the difference image. Then the statistical parameters of the difference image are estimated by Expectation Maximum algorithm assuming that the change and no-change pixel sets are modelled by Gaussian Mixture Model. As a result, a global initial threshold can be identified based on Bayesian decision theory. Next, a dynamic threshold operator is proposed by incorporating the membership value of each pixel generated by the Fuzzy c-means (FCM) algorithm and the global initial threshold. Lastly, the change map is obtained by segmenting the difference image utilizing the dynamic threshold proposed. Experimental results indicate that the proposed dynamic threshold method has significantly reduced the speckle noise comparing to the global threshold method. At the same time, weak change signals are detected and detail change information are preserved much better than the FCM does.     

Application of a homogenous patch mean kernel with within-class collaborative representation for hyperspectral imagery classification

Spectral–spatial-based classification methods demonstrate satisfying performance for hyperspectral imagery (HSI) classification. In this letter, in order to make full use of spectral and contexture information with simultaneously considering within-class information, we propose a new algorithm for HSI classification based on within-class collaborative representation and column generation (CG) strategy. The proposed accelerated homogeneous patch mean kernel (HPMK) can automatically assign a homogeneous patch for the target sample and represent the similarities between training set and assigned homogeneous patch in kernel feature space based on CG strategy. Further, for including intra-class information and improve classification efficiency, within-class collaborative representation classification (WCRC) is incorporated into new feature space to enhance the classification performance. Experiments on two real HSI data sets demonstrate that the proposed algorithm presents satisfying results in terms of classification accuracy and efficiency.     

Automated detection of impervious surfaces using night-time light and Landsat images based on an iterative classification framework

ABSTRACT

A novel workflow for automated detecting of impervious surface by using night-time light and Landsat images at the individual city scale is proposed. This approach is composed by of three steps. In the beginning, urban, peri-urban and rural regions are detected from the night-time light image by a contour line algorithm. Then, using Landsat TM image, region-specific spectral index analysis is employed to generate initial training samples of urban land covers. Finally, an iterative classification framework is applied to select new training samples by integrating spectral and spatial information and to obtain the final mapping result. Experimental results of two cities show that the proposed method produces higher classification accuracy than the ones using the manual-sampling methods. Moreover, further validations suggested that the spatial information is able to effectively increase the producer’s accuracy of impervious surface. This automated approach is potentially important for large-scale regional impervious surface mapping and application.     

Local histogram and discriminative learning-based hyperspectral data classification

A spectral–spatial hyperspectral image classification method is proposed in this article. The proposed method is local histogram and discriminative based classification (LHD). It is implemented in two steps. In the first step, a projection matrix is obtained by maximizing the class discrimination and preserving the local structure of data using both the spectral and spatial neighbours. In the second step, the contextual features are extracted from the local regions using the histogram as a nonparametric statistical estimate. Finally, with incorporating spectral and spatial features, the classification map is obtained. The experimental results demonstrate that the proposed method is superior to some state-of-the-art alternatives.     

Image segmentation using mean shift for extracting croplands from high-resolution remote sensing imagery

A new image segmentation algorithm based on mean shift (MS) is proposed, with an objective to single out croplands in high-resolution remote sensing imagery (HRRSI). The algorithm is composed of two parts. First, in order to improve the discontinuity preserving smoothing of MS filtering for cropland HRRSI, normal and uniform kernels are used to filter inner fields and boundary areas, respectively. A new spectral bandwidth estimation is also developed for better suppressing intra-field variation. Second, a two-stage region-merging technique, with the second stage combining mutual best-fit rule and iterative thresholding, is implemented. An HRRSI scene is used for validation, the results of which indicate good performance of our method.