Hyperspectral Remote Sensing of Urban Areas: An Overview of Techniques and Applications
01 Jan 2012-
TL;DR: In this article, the authors review the methods for urban classification using hyperspectral remote sensing data and their applications, and the general trends indicate that combined spatial-spectral and sensor fusion approaches are the most optimal for hyperspectra urban analysis.
Abstract: Over the past two decades, hyperspectral remote sensing from airborne and satellite systems has been used as a data source for numerous applications. Hyperspectral imaging is quickly moving into the mainstream of remote sensing and is being applied to remote sensing research studies. Hyperspectral remote sensing has great potential for analysing complex urban scenes. However, operational applications within urban environments are still limited, despite several studies that have explored the capabilities of hyperspectral data to map urban areas. In this paper, we review the methods for urban classification using hyperspectral remote sensing data and their applications. The general trends indicate that combined spatial-spectral and sensor fusion approaches are the most optimal for hyperspectral urban analysis. It is also clear that urban hyperspectral mapping is currently limited to airborne data, despite the availability of spaceborne hyperspectral systems. Possible future research directions are also discussed.
Citations
More filters
TL;DR: A new supervised classification algorithm for remotely sensed hyperspectral image (HSI) which integrates spectral and spatial information in a unified Bayesian framework and achieves better performance on one synthetic data set and two benchmark HSI data sets in a number of experimental settings.
Abstract: This paper presents a new supervised classification algorithm for remotely sensed hyperspectral image (HSI) which integrates spectral and spatial information in a unified Bayesian framework. First, we formulate the HSI classification problem from a Bayesian perspective. Then, we adopt a convolutional neural network (CNN) to learn the posterior class distributions using a patch-wise training strategy to better use the spatial information. Next, spatial information is further considered by placing a spatial smoothness prior on the labels. Finally, we iteratively update the CNN parameters using stochastic gradient decent and update the class labels of all pixel vectors using $\alpha $ -expansion min-cut-based algorithm. Compared with the other state-of-the-art methods, the classification method achieves better performance on one synthetic data set and two benchmark HSI data sets in a number of experimental settings.
257 citations
Cites background from "Hyperspectral Remote Sensing of Urb..."
...Since HSI provides detailed information on spectral and spatial distributions of distinct materials [48], it has been used for many applications, such as land-use [10], [33], [47] and landcover mapping [33], forest inventory [42], and urban area problems [52]....
[...]
TL;DR: Hyperspectral imaging is an optical method that provides a large amount of information about the investigated object that is useful for assessing tissue perfusion and its pathological conditions, making accurate surgical decisions, evaluating the health of dental structures, etc.
Abstract: Hyperspectral imaging is an optical method that provides a large amount of information about the investigated object. Its medical applications are reviewed in this article, including tumor delimitation and identification, assessing tissue perfusion and its pathological conditions (including some complications like diabetic foot ulceration), making accurate surgical decisions, evaluating the health of dental structures, etc. Many of the articles show very promising results that required brief comments by the authors. It is clear that choosing the appropriate hyperspectral imaging system for each medical field, together with the most reliable hyperspectral image processing methods, are the main goals of future studies, before hyperspectral imaging becomes a widely applicable evaluation method in medicine. The authors try to answer some questions on this topic and set up some directions for future research.
153 citations
Cites background from "Hyperspectral Remote Sensing of Urb..."
...Although originally developed for mining and geology, these systems are now used in fields such as agriculture (2), mineralogy (3), surveillance and target identification (4), astronomy (5), chemical imaging (6), and environmental studies (7)....
[...]
TL;DR: Wang et al. as mentioned in this paper proposed a new supervised classification algorithm for remotely sensed hyperspectral image (HSI) which integrates spectral and spatial information in a unified Bayesian framework. But, their method requires a large number of patches to be used.
Abstract: This paper presents a new supervised classification algorithm for remotely sensed hyperspectral image (HSI) which integrates spectral and spatial information in a unified Bayesian framework. First, we formulate the HSI classification problem from a Bayesian perspective. Then, we adopt a convolutional neural network (CNN) to learn the posterior class distributions using a patch-wise training strategy to better use the spatial information. Next, spatial information is further considered by placing a spatial smoothness prior on the labels. Finally, we iteratively update the CNN parameters using stochastic gradient decent (SGD) and update the class labels of all pixel vectors using an alpha-expansion min-cut-based algorithm. Compared with other state-of-the-art methods, the proposed classification method achieves better performance on one synthetic dataset and two benchmark HSI datasets in a number of experimental settings.
110 citations
TL;DR: A novel semi-supervised hyperspectral image classification framework is proposed which utilizes self-training to gradually assign highly confident pseudo labels to unlabeled samples by clustering and employs spatial constraints to regulate self- training process.
Abstract: Because there are many unlabeled samples in hyperspectral images and the cost of manual labeling is high, this paper adopts semi-supervised learning method to make full use of many unlabeled samples. In addition, those hyperspectral images contain much spectral information and the convolutional neural networks have great ability in representation learning. This paper proposes a novel semi-supervised hyperspectral image classification framework which utilizes self-training to gradually assign highly confident pseudo labels to unlabeled samples by clustering and employs spatial constraints to regulate self-training process. Spatial constraints are introduced to exploit the spatial consistency within the image to correct and re-assign the mistakenly classified pseudo labels. Through the process of self-training, the sample points of high confidence are gradually increase, and they are added to the corresponding semantic classes, which makes semantic constraints gradually enhanced. At the same time, the increase in high confidence pseudo labels also contributes to regional consistency within hyperspectral images, which highlights the role of spatial constraints and improves the HSIc efficiency. Extensive experiments in HSIc demonstrate the effectiveness, robustness, and high accuracy of our approach.
66 citations
TL;DR: This study was conducted to identify a technique that accurately maps impervious and pervious surfaces from WorldView-2 (WV-2) imagery to evaluate the capability of spectral-based classifiers to classify urban features.
Abstract: Urban areas consist of spectrally and spatially heterogeneous features. Advanced information extraction techniques are needed to handle high resolution imageries in providing detailed information for urban planning applications. This study was conducted to identify a technique that accurately maps impervious and pervious surfaces from WorldView-2 (WV-2) imagery. Supervised per-pixel classification algorithms including Maximum Likelihood and Support Vector Machine (SVM) were utilized to evaluate the capability of spectral-based classifiers to classify urban features. Object-oriented classification was performed using supervised SVM and fuzzy rule-based approach to add spatial and texture attributes to spectral information. Supervised object-oriented SVM achieved 82.80% overall accuracy which was the better accuracy compared to supervised per-pixel classifiers. Classification based on the proposed fuzzy rule-based system revealed satisfactory output compared to other classification techniques with an overal...
42 citations
Cites background from "Hyperspectral Remote Sensing of Urb..."
...…29, No. 3, 268–292, http://dx.doi.org/10.1080/10106049.2012.760006 2013 Taylor & Francis hyperspectral remote sensing is very effective for urban mapping (Taherzadeh & Shafri 2011); however, its high cost and limited coverage prohibit operation and frequent usage of such data (Shafri et al. 2012)....
[...]
...hyperspectral remote sensing is very effective for urban mapping (Taherzadeh & Shafri 2011); however, its high cost and limited coverage prohibit operation and frequent usage of such data (Shafri et al. 2012)....
[...]
References
More filters
4,943 citations
Book•
01 Jan 1978
TL;DR: This modern climatology textbook explains those climates formed near the ground in terms of the cycling of energy and mass through systems.
Abstract: This modern climatology textbook explains those climates formed near the ground in terms of the cycling of energy and mass through systems.
4,767 citations
"Hyperspectral Remote Sensing of Urb..." refers background in this paper
...Urban areas are characterised by a large variety of artificial and natural surface materials, influencing ecological (Arnold and Gibbons, 1996), climatic and energy (Oke, 1987) conditions....
[...]
TL;DR: This paper addresses the problem of the classification of hyperspectral remote sensing images by support vector machines by understanding and assessing the potentialities of SVM classifiers in hyperdimensional feature spaces and concludes that SVMs are a valid and effective alternative to conventional pattern recognition approaches.
Abstract: This paper addresses the problem of the classification of hyperspectral remote sensing images by support vector machines (SVMs) First, we propose a theoretical discussion and experimental analysis aimed at understanding and assessing the potentialities of SVM classifiers in hyperdimensional feature spaces Then, we assess the effectiveness of SVMs with respect to conventional feature-reduction-based approaches and their performances in hypersubspaces of various dimensionalities To sustain such an analysis, the performances of SVMs are compared with those of two other nonparametric classifiers (ie, radial basis function neural networks and the K-nearest neighbor classifier) Finally, we study the potentially critical issue of applying binary SVMs to multiclass problems in hyperspectral data In particular, four different multiclass strategies are analyzed and compared: the one-against-all, the one-against-one, and two hierarchical tree-based strategies Different performance indicators have been used to support our experimental studies in a detailed and accurate way, ie, the classification accuracy, the computational time, the stability to parameter setting, and the complexity of the multiclass architecture The results obtained on a real Airborne Visible/Infrared Imaging Spectroradiometer hyperspectral dataset allow to conclude that, whatever the multiclass strategy adopted, SVMs are a valid and effective alternative to conventional pattern recognition approaches (feature-reduction procedures combined with a classification method) for the classification of hyperspectral remote sensing data
3,607 citations
"Hyperspectral Remote Sensing of Urb..." refers methods in this paper
...Based on published work, the SVM seem to be the most effective method in the classification of hyperspectral data (Melgani and Bruzzone, 2004; Camps-Valls and Bruzzone, 2005; van der Linden et al., 2007; Fauvel et al., 2007; Waske et al., 2009; Misman et al., 2010; Shafri and Zeen, 2011)....
[...]
TL;DR: A wide range of strategies to reduce impervious surfaces and their impacts on water resources can be applied to community planning, site-level planning and design, and land use regulation as mentioned in this paper.
Abstract: Planners concerned with water resource protection in urbanizing areas must deal with the adverse impacts of polluted runoff. Impervious surface coverage is a quantifiable land-use indicator that correlates closely with these impacts. Once the role and distribution of impervious coverage are understood, a wide range of strategies to reduce impervious surfaces and their impacts on water resources can be applied to community planning, site-level planning and design, and land use regulation. These strategies complement many current trends in planning, zoning, and landscape design that go beyond water pollution concerns to address the quality of life in a community.
2,087 citations
"Hyperspectral Remote Sensing of Urb..." refers background in this paper
...Urban areas are characterised by a large variety of artificial and natural surface materials, influencing ecological (Arnold and Gibbons, 1996), climatic and energy (Oke, 1987) conditions....
[...]
TL;DR: A seminal view on recent advances in techniques for hyperspectral image processing, focusing on the design of techniques able to deal with the high-dimensional nature of the data, and to integrate the spa- tial and spectral information.
1,481 citations
"Hyperspectral Remote Sensing of Urb..." refers background in this paper
...In their study, Plaza et al. (2009) discussed, among others, the importance of SVM, morphological profiles, Markov Random Field (MRV) and hierarchical segmentation methods for urban mapping....
[...]
...A significant number of works involving hyperspectral data processing focus on exploiting the spectral component without incorporating the spatial information (Plaza et al., 2009)....
[...]