Mineral Identification Using Unsupervised Classification from Hyperspectral Data
01 Jan 2020-pp 259-268
TL;DR: K-means is unsupervised clustering algorithm used to classify the mineral and then further SVM is used to check the classification accuracy, and Principle component analysis (PCA) was used to reduce the dimension of data by band selection approach.
Abstract: Hyperspectral imagery is one of the research areas in the field of remote sensing. Hyperspectral sensors record reflectance of object or material or region across the electromagnetic spectrum. Mineral identification is an urban application in the field of remote sensing of Hyperspectral data. Challenges with the hyperspectral data include high dimensionality and size of the hyperspectral data. Principle component analysis (PCA) is used to reduce the dimension of data by band selection approach. Unsupervised classification technique is one of the hot research topics. Due to the unavailability of ground truth data, unsupervised algorithm is used to classify the minerals present in the remotely sensed hyperspectral data. K-means is unsupervised clustering algorithm used to classify the mineral and then further SVM is used to check the classification accuracy. K-means is applied to end member data only. SVM used k-means result as a labelled data and classify another set of dataset.
Citations
More filters
TL;DR: In this article, the authors analyzed the LULC changes during 1990-2018 as well as the growth and pattern of built-up surfaces in relation to the population growth and migration in the suburbs of Delhi metropolitan city which is also known as the National Capital Region (NCR).
93 citations
TL;DR: In this article, a small, unplanned growing up city, namely, Krishnanagar urban agglomeration (KUA), in India, to apply the prediction-adaptation-resilience (PAR) approach to analyze the future urban landscape resilience and sustainable development goals (SDGs).
40 citations
TL;DR: In this paper, the authors used remote sensing images with sub-meter spatial resolutions collected using the Worldview-2(WV-2) satellite to interpret mineralisationrelated geologic bodies and analyse the distribution relationships of mineralisation-controlling structures.
8 citations
TL;DR: In this article , support vector machine algorithms based on texture features and topographic features were used to identify the Lujavrite formation and the distribution of thermal anomalies associated with radioactive elements were inverted using Landsat 8 TIRS thermal infrared data.
Abstract: The harsh environment of high-latitude areas with large amounts of snow and ice cover makes it difficult to carry out full geological field surveys. Uranium resources are abundant within the Ilimaussaq Complex in the Narsaq region of Greenland, where the uranium ore body is strictly controlled by the Lujavrite formation, which is the main ore-bearing rock in the complex rock mass. Further, large aggregations of radioactive minerals appear as thermal anomalies on remote sensing thermal infrared imagery, which is indicative of deposits of highly radioactive elements. Using a weight-of-evidence analysis method that combines machine-learned lithological classification information with information on surface temperature thermal anomalies, the prediction of radioactive element-bearing deposits at high latitudes was carried out. Through the use of Worldview-2 (WV-2) remote sensing images, support vector machine algorithms based on texture features and topographic features were used to identify Lujavrite. In addition, the distribution of thermal anomalies associated with radioactive elements was inverted using Landsat 8 TIRS thermal infrared data. From the results, it was found that the overall accuracy of the SVM algorithm-based lithology mapping was 89.57%. The surface temperature thermal anomaly had a Spearman correlation coefficient of 0.63 with the total airborne measured uranium gamma radiation. The lithological classification information was integrated with surface temperature thermal anomalies and other multi-source remote sensing mineralization elements to calculate mineralization-favorable areas through a weight-of-evidence model, with high-value mineralization probability areas being spatially consistent with known mineralization areas. In conclusion, a multifaceted remote sensing information finding method, focusing on surface temperature thermal anomalies in high-latitude areas, provides guidance and has reference value for the exploration of potential mineralization areas for deposits containing radioactive elements.
3 citations
References
More filters
TL;DR: The challenges of using deep learning for remote-sensing data analysis are analyzed, recent advances are reviewed, and resources are provided that hope will make deep learning in remote sensing seem ridiculously simple.
Abstract: Central to the looming paradigm shift toward data-intensive science, machine-learning techniques are becoming increasingly important. In particular, deep learning has proven to be both a major breakthrough and an extremely powerful tool in many fields. Shall we embrace deep learning as the key to everything? Or should we resist a black-box solution? These are controversial issues within the remote-sensing community. In this article, we analyze the challenges of using deep learning for remote-sensing data analysis, review recent advances, and provide resources we hope will make deep learning in remote sensing seem ridiculously simple. More importantly, we encourage remote-sensing scientists to bring their expertise into deep learning and use it as an implicit general model to tackle unprecedented, large-scale, influential challenges, such as climate change and urbanization.
2,095 citations
TL;DR: In this article, the authors analyze the challenges of using deep learning for remote sensing data analysis, review the recent advances, and provide resources to make deep learning in remote sensing ridiculously simple to start with.
Abstract: Standing at the paradigm shift towards data-intensive science, machine learning techniques are becoming increasingly important. In particular, as a major breakthrough in the field, deep learning has proven as an extremely powerful tool in many fields. Shall we embrace deep learning as the key to all? Or, should we resist a 'black-box' solution? There are controversial opinions in the remote sensing community. In this article, we analyze the challenges of using deep learning for remote sensing data analysis, review the recent advances, and provide resources to make deep learning in remote sensing ridiculously simple to start with. More importantly, we advocate remote sensing scientists to bring their expertise into deep learning, and use it as an implicit general model to tackle unprecedented large-scale influential challenges, such as climate change and urbanization.
629 citations
TL;DR: A novel network architecture, fully Conv–Deconv network for unsupervised spectral–spatial feature learning of hyperspectral images, which is able to be trained in an end-to-end manner and an in-depth investigation of learned features is introduced.
Abstract: Supervised approaches classify input data using a set of representative samples for each class, known as training samples . The collection of such samples is expensive and time demanding. Hence, unsupervised feature learning, which has a quick access to arbitrary amounts of unlabeled data, is conceptually of high interest. In this paper, we propose a novel network architecture, fully Conv–Deconv network, for unsupervised spectral–spatial feature learning of hyperspectral images, which is able to be trained in an end-to-end manner. Specifically, our network is based on the so-called encoder–decoder paradigm, i.e., the input 3-D hyperspectral patch is first transformed into a typically lower dimensional space via a convolutional subnetwork (encoder), and then expanded to reproduce the initial data by a deconvolutional subnetwork (decoder). However, during the experiment, we found that such a network is not easy to be optimized. To address this problem, we refine the proposed network architecture by incorporating: 1) residual learning and 2) a new unpooling operation that can use memorized max-pooling indexes. Moreover, to understand the “black box,” we make an in-depth study of the learned feature maps in the experimental analysis. A very interesting discovery is that some specific “neurons” in the first residual block of the proposed network own good description power for semantic visual patterns in the object level, which provide an opportunity to achieve “free” object detection. This paper, for the first time in the remote sensing community, proposes an end-to-end fully Conv–Deconv network for unsupervised spectral–spatial feature learning. Moreover, this paper also introduces an in-depth investigation of learned features. Experimental results on two widely used hyperspectral data, Indian Pines and Pavia University, demonstrate competitive performance obtained by the proposed methodology compared with other studied approaches.
234 citations
TL;DR: Two methods, based on the concept of spectral unmixing and unsuper supervised classification, are proposed to obtain thematic maps at a finer spatial scale in a totally unsupervised way and clearly show the comparative effectiveness of the proposed method with respect to traditionalunsupervised methods.
64 citations
TL;DR: It is shown that the difference between the global diversity of clusters and the sum of each cluster’s local diversity of their members can be used as an effective indicator of the optimality of the number of clusters, where the diversity is measured by Rao's quadratic entropy.
Abstract: It is an important and challenging problem in unsupervised learning to estimate the number of clusters in a dataset. Knowing the number of clusters is a prerequisite for many commonly used clustering algorithms such as \textit{k}-means. In this paper, we propose a novel diversity based approach to this problem. Specifically, we show that the difference between the global diversity of clusters and the sum of each cluster’s local diversity of their members can be used as an effective indicator of the optimality of the number of clusters, where the diversity is measured by Rao’s quadratic entropy. A notable advantage of our proposed method is that it encourages balanced clustering by taking into account both the sizes of clusters and the distances between clusters. In other words, it is less prone to very small “outlier” clusters than existing methods. Our extensive experiments on both synthetic and real-world datasets (with known ground-truth clustering) have demonstrated that our proposed method is robust for clusters of different sizes, variances, and shapes, and it is more accurate than existing methods (including elbow, Calinski-Harabasz, silhouette, and gap-statistic) in terms of finding out the optimal number of clusters.
45 citations