Multitask learning is a widely recognized technique in the field of computer vision and deep learning domain. However, it is still a research question in remote sensing, particularly for hyperspectral imaging (HSI). Moreover, most of the research in the remote sensing domain focuses on small and single-task-based annotated datasets, which limits the generalizability and scalability of the developed models to more diverse and complex real-world scenarios. Thus, in this study, we propose a multitask deep learning model designed to perform multiple classification and regression tasks simultaneously on hyperspectral images. We validated our approach on a large hyperspectral dataset called TAIGA, which contains 13 forest variables, including three categorical variables and ten continuous variables with different biophysical parameters. We design a sharing encoder and task-specific decoder network to streamline feature learning while allowing each task-specific decoder to focus on the unique aspects of its respective task. In addition, a dense atrous pyramid pooling layer and attention network were integrated to extract multiscale contextual information and enable selective information processing by prioritizing task-specific features. Furthermore, we computed multitask loss and optimized its parameters for the proposed framework to improve the model performance and efficiency across diverse tasks. A comprehensive qualitative and quantitative analysis of the results shows that the proposed method significantly outperforms other state-of-the-art methods. We trained our model across ten seeds/trials to ensure robustness. Our proposed model demonstrates higher mean performance while maintaining lower or equivalent variability. To make the work reproducible, the codes will be available athttps://github.com/Koushikey4596/Multitask-Deep-Learning-Model-for-Taiga-datatset.

Hyperspectral imaging is essential for the detailed and accurate identification of materials and features across various applications, including environmental monitoring and agricultural assessment. However, processing large-scale hyperspectral images is time-consuming and requires huge computational resources attributed to their volume. Additionally, classification methods lacking spatial information and focusing only on spectral information are inadequate with the feature complexity of the dataset. Hence, to overcome the above-mentioned challenges, there is a dire need for advanced algorithms capable of efficiently processing large-scale hyperspectral data. This paper highlights the motivation and need for multitask learning models for hyperspectral datasets. A large-scale hyperspectral dataset called TAIGA, which comprises both categorical and continuous forest variables, is considered. The proposed deep learning model addresses data imbalance and loss function concerns by incorporating spectral as well as spatial information to develop a multitask model. This work emphasizes the importance of accounting for correlations between data and tasks when designing a relevant model. As a result, the study leads to more efficient model training, reduced data requirements, and potentially improved overall predictive capabilities. We achieved overall accuracy for categorical variables as high as 98.25% and mean absolute error as low as 0.019 for continuous variables.

Texture feature extraction is very important in landuse land cover(LULC) classification of satellite images. Through this paper, a new method for texture feature extraction is presented which uses singular value decomposition(SVD) and gray level cooccurance matrix (GLCM). Here we test for capabilities of the singular values thus generated for classification of textures. We also try to find out if these singular values can be used as a substitute for Haralick texture features. In this proposed method sample images are multi-thresholded to reduce the dimensionality. GLCM is generated for each image-patch after applying the thresholds. Later, the SVD decomposition of GLCM provides singular values which are used as a feature vector for classification of the image patches. We evaluated our proposed technique based on three criteria a) images from totally different classes b) images from same class but with different textures c) images from same class, same texture but different orientation. We classified the images using minimum distance to mean(MDM), SVM using radial bias kernel (cSVM) and SVM using quantum kernel (qSVM). We used IBM gate-based qiskit to generate a quantum kernel and classify using qSVM. From the experiment we conclude that singluar values are good and stable as textures features and greatly enhance the texture classification. Singular Values along with quantum kernel have produced very promising results.

Recently, Convolutional Neural Network (CNN) has been widely used for the classification of hyperspectral images. However, with the availability of limited training sample data in hyperspectral scenes, classification performance is majorly affected. In this research work, we propose a shuffled dilated CNN-based classification method for hyperspectral classification using a transfer learning approach. The proposed model consists of a dilated convolution layer to provide a larger receptive field and a shuffled block to enhance the connection between different layers. This helps in the development of a classification model having fewer parameters with better efficiency. To evaluate the performance, hyperspectral datasets have been considered from the same sensor as well as cross-sensor with similar spectral and spatial features. We observed an improvement of 4% overall accuracy using the transfer learning approach compared to the classification results without using the transfer learning approach. The experimental results demonstrate the effectiveness of using the proposed methodology with a transfer learning approach for the classification of hyperspectral data with a limited number of labeled training samples.

Quantum Machine Learning (QML) is a new interdisciplinary branch that combines quantum computing with machine learning. It is emerging as an alternative to classical machine learning which exploits the quantum mechanical properties of entanglement and superposition to express the hidden patterns in the data. This reduces computational resources as well as the time required for processing. In this study, we tried to address the challenge of landuse land cover classification for classes with similar spectral signatures and small texture dissimilarity. Totally 6 textural features are extracted from a multispectral image using grey level co-occurrence matrix applied on 1st PC component to classify images into the sub-classes(residential, highway, and industrial) of the main built-up class. We used IBM gate-based qiskit to generate a quantum kernel and classify the images using QSVC. Quantum kernels are very expressive when compared to their classical counterparts and can learn complex data more efficiently. The overall accuracy of classification by QSVC is comparable to that of the classical SVC. We summarize our results by saying that QSVC performs better than SVC.

Salient object detection using hyperspectral images is crucial for various image processing and computer vision applications. Many studies considering spectral information have been developed, extracting only low-level features from a hy-perspectral image. In this research work, a dataset specifically developed for salient object detection called HS-SOD is considered exploiting both spatial and spectral information equally. To include spatial information, Extended Morpho-logical Profile (EMP) has been considered. EMP incorpo-rates spatial characteristics by including nearby pixel information. A convolution neural network (CNN) is integrated with extended morphology to extract high-level features. It detect objects of multiple spatial scales and ratios, preserving boundary edges. We observed an improvement of 5 % in overall accuracy while using EMP with CNN compared to that of using EMP without CNN. Thus, the experimental re-sults demonstrate the effectiveness of EMP with CNN on the hyperspectral datasets.

Salient object detection in hyperspectral images is of interest in various image processing and computer vision applications. Many studies considering spectral information have been reported, extracting only low-level features from a hyperspectral image. This paper proposes a Convolutional Neural Network (CNN) based salient object detection method using hyperspectral imagery to utilise spatial and spectral information simultaneously. The proposed methodology incorporates Extended Morphological Profile (EMP) followed by a CNN to utilise the information from nearby pixels and high-level features simultaneously. We have evaluated the performance of the proposed approach on two independent datasets to verify the generalisation ability, viz. 1) Hyperspectral Salient Object Detection Dataset (HS-SOD) and 2) Pavia University dataset. An extensive quantitative analysis of the results revealed that the proposed method significantly outperforms other state-of-the-art methods by approximately ≥ 2% of AUC (Area Under receiver operating characteristic Curve) and F-measure and lower mean absolute error for both datasets.

Hyperspectral Imagery (and Remote Sensing in general) captured from UAVs or satellites are highly voluminous in nature due to the large spatial extent and wavelengths captured by them. Since analyzing these images requires a huge amount of computational time and power, various dimensionality reduction techniques have been used for feature reduction. Some popular techniques among these falter when applied to Hyperspectral Imagery due to the famed curse of dimensionality. In this paper, we propose a novel approach, LEt-SNE, which combines graph based algorithms like t-SNE and Laplacian Eigenmaps into a model parameterized by a shallow feed forward network. We introduce a new term, Compression Factor, that enables our method to combat the curse of dimensionality. The proposed algorithm is suitable for manifold visualization and sample clustering with labelled or unlabelled data. We demonstrate that our method is competitive with current state-of-the-art methods on hyperspectral remote sensing datasets in public domain.

Remotely sensed images are suitable resources to examine the extent and impact of fire on the land-scape. This paper presents a multiresolution approach to detect changes due to fire. A large number of change detection methods utilizing wavelet features have been developed. However, usage of contourlets is limited in context of change detection application. The aim of this paper is to investigate and propose a texture based model utilizing contourlets in parallel-processing framework to detect changes due to fire. To achieve near real time processing, the proposed algorithm is implemented on NVIDIA's Graphical Processing Unit. The results are compared with the changes detected by different thresholding methods in the feature space. It is found that the proposed method exhibits high change detection accuracy with better edge continuity. The results show that the GPU implementation improves the performance in terms of execution time as compared to the sequential implementation.

Single-image super-resolution (SISR) techniques attempt to reconstruct the finer resolution version of a given image from its coarser version. In the SISR of hyperspectral data sets, the simultaneous consideration of spectral bands is crucial for ensuring the spectral fidelity. However, the high spectral resolution of these data sets affects the performance of conventional approaches. This research proposes the design of 3-D convolutional neural network (CNN)-based SISR architectures that can map the spatial-spectral characteristics of hypercubes to a finer spatial resolution. The proposed approaches facilitate the simultaneous optimization of sparse codes and dictionaries with regard to the super-resolution objective. Our main hypothesis is that the consideration of spectral aspects is essential for the spatial enhancement of hyperspectral images. Also, we propose that the regularized deconvolution of a coarser-scale hypercube, using learned 3-D filters, yields the required high-resolution version. Based on these hypotheses, a convolution-deconvolution framework is proposed to super-resolve the hypercubes in parallel with the reconstruction of a set of regularizing features. Novel sparse code optimization sub-networks proposed in this article give better performance than the existing strategies. The endmember similarities and hyperspectral image prior are considered while designing the proposed loss functions. In order to improve the generalizability, a collaborative spectral unmixing strategy is employed to refine the spectral base of the super-resolved result. The spatial-spectral accuracy of the super-resolved hypercubes, in terms of the validity of regularizing features and endmembers, is explored to devise an optimal ensemble strategy. The experiments, over different data sets, confirm better accuracy of the proposed frameworks compared to the prominent approaches.

The objective of secure outsourcing is to enforce authorizations specified by the data owner on the outsourced data as service provider cannot be trusted. It protects the location data from attackers while allowing authorized users to issue spatial queries that are executed efficiently by service provider. In this paper, we are proposing secure outsourcing mechanism for geospatial vector data by utilizing Hilbert R tree for transformation and indexing of outsourced data and Merkle Hash tree for secure storage of outsourced data as well as for preserving integrity of the query results generated by service provider.

Identifying informal settlements might be one of the most challenging tasks within urban remote sensing in developing countries. Despite several studies, the task of identifying informal settlements remains a challenge. Formal settlements are mapped sufficiently with distinct features, however this does not hold for informal settlements because of their microstructure and instability of shape, the detection of these settlements is substantially more challenging. Therefore, more advanced multi-scale methods of image analysis are necessary, which act as a spatial basis for informal settlement identification. This paper proposes a texture based scheme using contourlet based multiresolution method to capture anisotropic features for informal settlement area identification. The approach consists of three main steps: identifying area of interest, extraction of the most discriminative texture features and creation of a classifier that automatically identifies the informal settlements. The algorithm is extensively tested and results are compared with wavelet based and grey level co-occurrence based standard texture classification methods. A comparative analysis is carried out using minimum distance to mean in transformed feature space. The performance is evaluated in terms of sensitivity, specificity, precision and overall accuracy. It is found that the proposed method shows better class-discriminating power as compared to existing methods and overall classification accuracy of 93-96%.

Dense stereo image matching in remotely sensed images is a challenging problem, though it has been studied for more than two decades, due to occlusions, discontinuities, geometric, and radiometric distortions. A novel multiobjective fitness function-based dense stereo matching approach using genetic algorithms (GAs) is proposed in this paper. The proposed method is useful for estimating dense disparity map with an improved number of inliers for a stereo image pair, despite the constraint of finding correct disparity at depth discontinuities. In this paper, the steps of GA, such as initialization of the population, fitness function, and crossover and mutation operation, are designed and implemented to effectively deal with the problem of dense stereo image matching. To initialize the population, a Scale Invariant Feature Transform (SIFT) descriptor is computed for each pixel and multiple-size window-based matching is performed, using the similarity measures: 1) Euclidean distance and 2) spectral angle mapper. The generated disparity maps are pruned to choose a suitable subset using the designed fitness functions, considering the constraints related to stereo image pair, such as epipolar constraint, which encodes the epipolar geometry and the similarity measure that is useful to decide accuracy of the correspondences. The two objective functions are the number of inliers computed using the fundamental matrix and an energy minimization function, considering discontinuities and occlusions. The usefulness of this approach for remotely sensed stereo image pairs is demonstrated by improving the number of inliers and favorably comparing with state-of-the-art dense stereo image matching methods.

This paper presents a multiresolution textural approach to change detection in multi-temporal synthetic aperture radar (SAR) images. The proposed approach exploits curvelet and contourlet based multi-scale transforms for SAR data where textural information is extracted at various scales and in different directions in terms of statistical moments and energy to generate the feature vectors. The L1-norm is used to generate the difference image, which is thresholded using the maximum entropy principle to obtain final change detection map. The results are compared with the changes detected by wavelet based textural features. Accuracy assessment is performed for change maps and comparative analysis is carried out in terms of missed changes, false-alarms and overall accuracies. It is found that the proposed method exhibits high change detection accuracy with better edge continuity compared to wavelet based methods.

In this paper, we propose the use of network inversion for modeling the variation of class distributions with scale. Unlike the state of the art methods that predict the mapping between coarser and finer scale patches without considering the distributions at coarser scale, our approach uses coarser scale features for effective reconstruction. This is the pioneer work of using network inversion for the purpose. Analysis over the proposed framework reveals that both the computational performance and accuracy varies with the depth of the network as well as the size and number of filters in each layer. Also the performance of the approach has been found to improve with the increase in the number of input feature maps. Investigations over standard datasets indicate that the proposed approach performs much better than the recent sub-pixel classification as well as super resolution techniques.

Independence of Synthetic Aperture Radar (SAR) images from weather and sun illumination helps us in maritime surveillance like an oil spill, to identify the illegal fishery activity and unauthorized marine vessels. Detection of marine vessels is like a point target detection in coarse to moderate resolution images. Detection of point targets is adversely affects by speckle noise. Object detection in SAR images, without explicitly reducing the speckle noise is one of the challenging task. In this paper two algorithms are presented, which show how improvement in the power of point target lead us to reduced number of false alarms. The first algorithm has three parts. First part uses Grave matrix, which is a ( 2×2) Hermitian power matrix, generated by the multiplication of Sinclair matrix and conjugate of Sinclair matrix. Second part uses discrimination criteria to discriminate between clutter and vessels based on eigenvalue of the Grave matrix (G). The third part, fill the gaps by using morphological dilation. The only difference between the first and the second algorithms is that, in the second algorithm we uses Sinclair matrix (S2) instead of Grave matrix. Both the algorithms tested on two full-polarization (HH, HV, VH, VV) datasets and the results show the importance of scattering power matrix (G) as compared to scattering matrix (S2) for point target detection. The first data set is of size 498*498 captured by AlOS-1 PALSAR L-band data, covering the coastal region of Singapore. The second data set is of size 472*472, covering the coast of Vancouver acquired in C-band by Radarsat-2.

It is found very often that training data contains unequal number of representative samples for classes. Some of the classes might be represented by a larger number of samples while the rest with lower number of samples. Classification of remote sensing images with imbalanced class distribution could result in a significant drawback in the classification performance attainable by most standard classifier learning algorithms which assume a relatively balanced class distribution and equal misclassification costs. So it is worth exploring if ensemble method could give an improved performance under the condition of imbalanced training data. In the proposed work, Support Vector Machine (SVM) is used as base classifiers in the ensemble committee. An ensemble of SVMs will be constructed using popular Bagging method. Standard Hyperspectral data such as Salinas is used as test data. The proposed work will explore the efficiency of ensemble technique in improving classification accuracy, even in cases of robust classifier such as SVM.

Hyperspectral images are rich in spectral content but their spatial resolution is relatively poor. It can lead to mixed pixels and sub-pixel targets. In order to improve the reliability of information provided by hyperspectral image analysis and make the results practically usable, one needs to improve their spatial resolution. Due to physical constraints and associated cost, increasing the resolution by improving the sensors may not be a practical option. Thus one effective solution is some form of post-processing of hyperspectral data. Such an algorithmic resolution enhancement is called “super-resolution”. In this paper single image super-resolution of hyperspectral image has been attempted. The use of Hopfield Neural Network for successful landuse/landcover classification of Hyperspectral image has been shown. A successful attempt was made to improve initialization of the Hopfield neural network. The results were verified visually as well as statistically.

The monitoring of the water cycle at the Earth surface which tightly interacts with the climate change processes as well as a number of practical applications (agriculture, soil and water quality assessment, irrigation and water resource management, etc…) requires surface temperature measurements at local scale. Such is the goal of the Indian-French high spatio-temporal TRISHNA mission (Thermal infraRed Imaging Satellite for High-resolution Natural resource Assessment). The scientific objectives of the mission and research work conducted to consolidate the mission specifications are presented. Progress in modelling of surface fluxes is then discussed. The main specifications of the mission such as the revisit, the spatial resolution, the overpass time, the spectral bands and the orbit are analyzed and justified. The resulting baseline of the mission is given.

Hyperspectral images captured by airborne sensors provide huge information about earth cover. Human eye can see reflectance from limited wavelengths whereas hyperspectral images contain more than 200 bands captured in visible and infrared region of light. Classification is one of the methods used to analyse this information for decision making. Several techniques are available in the literature for supervised, unsupervised, and semi-supervised classification of hyperspectral images. Training samples for supervised classification are generated by ground sample collection from the actual site captured in the image. These training samples play an important role in classification and can greatly influence classification accuracy. In recent research, for selecting training samples `pixel purity' based new methods were developed. However, in some cases, it might not be possible to gather ground truth of selected pixel. This paper advocate for a novel method inspired by Ant Colony System for selection of a subset from already developed training samples to achieve better classification results. It is interesting to see that a small subset of existing training samples achieve equivalent or in some cases better accuracy when compared to the use of all available training samples.

This letter investigates the use of coarse-image features for predicting class labels at a given finer spatial scale. In this regard, two unsupervised subpixel mapping approaches, a semivariogram method, and a pixel-affinity based method are proposed. Furthermore, segmentation-based spectral unmixing is explored so as to address the spectral variability and nonconvexity of classes. In addition, the gradient information is employed to resolve uncertainties in the unmixing process. The proposed modifications based on pixel-affinity and semivariogram have produced an accuracy improvement of 5% or more over the state-of-the-art approaches.

AdaBoost is a popular ensemble method utilized in pattern recognition problems that are considered tough. Besides being a robust technique it does suffer from few limitations viz. size of training data and presence of noise in training data. In this context, we proposed a novel technique called Perspective Based Model (PBM) for ensemble creation in case of multispectral data analysis. In the present paper, we evaluate its performance in terms of classification accuracy against AdaBoost.M2. Preliminary results show higher accuracy through PBM compared to a single classifier and promising classification results for PBM compared to AdaBoost.M2.

Classification of hyperspectral satellite images is one of the methods of extracting information which further can be used in the form of thematic maps or feature extraction. In order to improve classification accuracy along with spectral classifiers some spatial information driven preprocessing or postprocessing techniques are used. Here, we have proposed joint spatial-spectral information based classifier using Ant Colony Optimization which removes requirement of computationally hard sequential filtering. Proposed method has achieved an improvement of 5-10% in classification accuracy over traditional methods like Support Vector Machine, ANN, and SAM.

Geographical Information System (GIS) is used to collect, manipulate, analyze, and display the geospatial data. The compilation and management of this spatial data is expensive and time consuming task. Due to rapid growth of distributed networks and Internet, it becomes easy to handle data but at the same time it becomes easy to copy or distribute the spatial data. Therefore copyright protection, authenticity, and spatial data source tracing have become important issues. One of the remarkable methods to solve these issues is digital watermarking. The objective of this paper is to propose a digital watermarking scheme which can be used for copyright protection of geospatial data using vector data as a watermark. In the proposed algorithm, vector watermark is embedded in low frequency coefficients at third level of wavelet decomposition. Experimental results show that the proposed algorithm is robust against all the attacks except compression attack for using vector data as watermark. It has been observed that increase in strength of embedding increases visual degradation.

Multi-resolution analysis (MRA) has been successfully used in image processing with the recent emergence of applications to texture classification. Several studies have investigated the discriminating power of wavelet-based features in various applications such as image compression, image denoising, and classification of natural textures. Recently, the curvelet and contourlet transforms have emerged as new multi-resolution analysis tools to deal with non-linear singularities present in the image. This article explores and proposes a texture based classification of remotely sensed multispectral images using features derived from the wavelet, curvelet and contourlet transforms. These features characterize the textural properties of the images and are used to train the classifier to recognize each texture class. Using these MRA based feature descriptors class separability is defined in feature space. The results are compared with Grey Level Co-occurrence Matrix (GLCM) based statistical features.

In ensemble methods applied to base classifiers that generate class probabilities, such classification outputs from individual ensemble members are combined using soft voting method. Weights are computed using an optimization function that is based upon the test performance of all trained ensemble members. However, in special case such as bagging method where samples are chosen randomly, there is chance of repetition of same samples, lacking of class-specific global representative samples which could influence the behavior of members. Aforementioned a priori information on representative samples could be used to reinforce or dilute weightage values for a particular class. The present paper is an attempt to explore the possibility of improving class-specific weighted soft voting using proportionality information derived from repetition and intra-class variability (global representation) of training samples. Additional class-specific weights are assigned in the computation of final decision. Preliminary results show slight improvement in accuracies.

Spectral unmixing techniques decompose the pixels into constituent fractions in order to extract the subpixel information. This study reviews spectral unmixing techniques from a perspective different from earlier approaches in that the problem is studied from a classification as well as clustering perspective. In this research, we focus on addressing some core issues of spectral unmixing such as endmember variability, requirement of pure endmember values, and initialization sensitivity modelling. We propose a Support Vector Machine (SVM) based unmixing technique that incorporates endmember spectral variability. The method uses endmember extraction techniques to give optimal performance even in the absence of training samples. Further, our study presents an alternation of FCM based method for incorporating spectral variability, and the approach is found to be resilient to the brightness variation. An automatic approach for fuzziness parameter selection is also introduced. The sensitivity of FCM towards endmember initialization has been considerably reduced by optimizing the initial seed selection. The proposed approaches have been analyzed over various standard datasets.

Super-resolution techniques use subpixel information to predict high resolution classification maps from coarse images. This study investigates for an unsupervised super-resolution approach which considers the image features to predict target spatial dependencies. Novelty of the approach is that the convolution neural networks and deep autoencoders are explored in this context. Evaluation over standard datasets revealed that the proposed method is more effective than the state of art unsupervised approaches. The method is also found to be preferable over variogram based approaches for complex scenes. This study also compares the effectiveness of shallow and deep networks and investigates the possible assessment of the optimal depth for the learning network. This technique can be further extended to a supervised framework.

We propose a novel coclustering-based domain-adaptation algorithm for simultaneously generating classification maps for a set of remote sensing (RS) multitemporal images in this letter. Unsupervised domain-adaptation techniques consider two different but related domains: a source domain with ample number of labeled samples and a target domain with no labeled data. The task at hand is to build an inference model exploring the available data that is expected to work consistently well in both the domains. This is a challenging problem, since the probability distributions governing both the domains are substantially different leading to the violation of the probably approximate correct assumptions of statistical learning theory. We consider an even complex scenario in this letter by assuming the absence of source domain training samples in the learning process. Our algorithm broadly consists of two stages: first, data from both the domains are projected into a common subspace using geodesic flow kernel in a Grassmannian manifold, and we further propose an iterative coclustering technique to obtain the consistent clustering outcomes for both the domains in the newly defined space. In line with the traditional domain-adaptation approaches, we also consider that both the domains contain the same set of semantic land-cover classes. The proposed method is simple, scalable, and results in highly precise clustering outputs for standard RS data sets.

This article presents a hybrid technique for noise filtering of remotely sensed images based on multiresolution analysis (MRA). Multiresolution techniques provide a coarse-to-fine and scale-invariant decomposition of images for image interpretation. Further, noise being one of the biggest problems in image analysis and interpretation for further processing, is effectively handled by multiresolution methods. The paper proposes a hybrid scheme based on wavelet and curvelet transforms on high resolution multispectral images acquired by the Quickbird and medium resolution Landsat Thematic Mapper satellite systems. By comparative analysis, the hybrid approach of curvelet and wavelet for heterogeneous and homogeneous areas has proved to be better than the others. Results are illustrated using Quickbird and Landsat images for proposed method and compared with wavelets and curvelet based noise filtering.

In the present paper, efficiency and competence of an ensemble method is explored in the context of large number of available spectral information. Classification results of ensemble method are compared with the results generated by a single classifier utilizing all spectral channels. In the present study, an ensemble committee is constructed by distributing spectral channels among five members of the committee to satisfy diversity criteria. Each spectral channel is representative of a particular wavelength and each Earth feature has its own spectral signature to a specific wavelength. Taking advantage of this fact, the present study attempts to explore the possibility of constructing diverse ensemble members in addition to achieving improved classification accuracy with respect to hard-to-discriminate image objects. Classification results obtained are promising. Overall classification accuracy is better through ensemble method. Some hard-to-discriminate objects are correctly identified. However, in some cases we obtained mixed results.

This paper addresses the problem of land-cover classification of remotely sensed image pairs in the context of domain adaptation. The primary assumption of the proposed method is that the training data are available only for one of the images (source domain), whereas for the other image (target domain), no labeled data are available. No assumption is made here on the number and the statistical properties of the land-cover classes that, in turn, may vary from one domain to the other. The only constraint is that at least one land-cover class is shared by the two domains. Under these assumptions, a novel graph theoretic cross-domain cluster mapping algorithm is proposed to detect efficiently the set of land-cover classes which are common to both domains as well as the additional or missing classes in the target domain image. An interdomain graph is introduced, which contains all of the class information of both images, and subsequently, an efficient subgraph-matching algorithm is proposed to highlight the changes between them. The proposed cluster mapping algorithm initially clusters the target domain data into an optimal number of groups given the available source domain training samples. To this end, a method based on information theory and a kernel-based clustering algorithm is proposed. Considering the fact that the spectral signature of land-cover classes may overlap significantly, a postprocessing step is applied to refine the classification map produced by the clustering algorithm. Two multispectral data sets with medium and very high geometrical resolution and one hyperspectral data set are considered to evaluate the robustness of the proposed technique. Two of the data sets consist of multitemporal image pairs, while the remaining one contains images of spatially disjoint geographical areas. The experiments confirm the effectiveness of the proposed framework in different complex scenarios.

A novel ant colony optimization based domain adaptation for satellite images has been proposed in the paper. Given a source domain and a target domain image, it has been considered here that we have labeled training data for the source domain image. The goal is to classify the target domain image for which no prior information is available. The proposed method exploits the advantages of ant colony optimization for performing the adaptation process. The target domain data is first over-clustered and a cross-domain cluster matching strategy based on ant movement is followed next to match the target domain clusters to the source domain land-cover classes. Experiments suggest good matching capabilities of the proposed technique.

Edge detection in SAR images has always been a challenge due to the effects of random interference of coherent signal. Unlike optical images, boundary delineation of regions is relatively ineffective for SAR images. The usual edge detectors, successful with incoherent images, yield poor results when applied to radar images, especially those with a small number of looks. Radon spectrum can be used to find the local orientations in an image which embed the edge information [1]. In this paper, we demonstrate the capability of Radon transform to efficiently extract edges from a SAR image covering agricultural landform. One approach to edge detection in polarimetric SAR images is to perform the edge detection separately for each of the polarization channels and subsequently combine the results using a fusion operation [2] [3]. We have made use of HH, HV and VV channels and fused the edge maps of these individual bands with Boolean ‘AND’ operator. This type of fusion has preserved the prominent edge information reasonably while suppressing spurious ones.

Segmentation of satellite images using a novel adaptive non parametric mean-shift clustering algorithm is proposed in this paper. Image segmentation refers to the process of splitting up an image into its constituent objects. It is also an important step in bridging the semantic gap between low level image interpretation and high level visual analysis. Mean-shift technique is based on the concept of kernel density estimation. It has been applied successfully in diverse vision related tasks including segmentation. The performance of the mean shift algorithm is greatly affected by the size of the parzen window and the terminating criteria. These two issues have been taken care of here in a purely statistical framework. The efficiency of this newly developed adaptive clustering has been judged for segmentation of any initially oversegmented satellite image. The notion of object based image analysis is preserved by initially over segmenting the image by watershed technique. Extensive experiments on several multispectral satellite images have confirmed the effectivity of this proposed approach in comparison to some widely used state of the art segmentation methods.

Many countries are producing high quality remotely sensed images from spaceborne sensors mounted on Earth orbiting satellites. One of the handicaps in the spread of this technology among endusers is lack of trained manpower and sometimes lack of resources for teaching and imparting training in this area. In this paper, a system developed to address this gap is described where an e-Tutor combined with a virtual laboratory offer educational content, demonstrations and exercises, case studies, self-assessment quizzes and questions, instructor's resources and lastly a self-contained virtual laboratory for hands-on sessions. The content is easily adequate for one semester intensive course on digital image processing for remote sensing or two semesters if covered at a leisurely pace.

Desert regions are typically characterized by the texture of sand that particular place and the orientation of dunes. We have attempted to visualize and quantify the sand dunes by local surface orientations calculated by Radon transform of sliding window region that covers entire image. This kind of representation can be used as a pre-processing step for segmentation and change detection of sand dunes. Hough transform, though traditionally used [1] to extract linear features is not very suitable for natural images. Radon transform on the other hand is robust to noise [2][3] comparatively and can effectively be used to find dune orientations. First of all an edge image is generated using canny operator or morphological operators and then local orientations are computed. Populations residing near desert land scapes have Livestock and marginal cultivation. To protect their interest's administrative setup need to have accurate understanding of how the directions of sand dunes are changing with respect to time. It is seen that for desert environments, remotely sensed satellite imagery can reliably be used, to identify regions that are nontextured and with various dune orientations using radon transform.

Image classification is an important task for many aspects of global change studies and environmental applications. This paper emphasizes on the analysis and usage of different advanced image classification techniques like Cloud Basis Functions (CBFs) Neural Networks, Artificial Neural Networks (ANN) and Support Vector Machines (SVM) for object based classification to get better accuracy. For comparison, adaptive Gaussian filtered images were classified using ANN and post-processed using relaxation labeling process (RLP). The results are demonstrated using high spatial resolution remotely sensed images.