Pub Date : 2012-12-31DOI: 10.1109/PPRS.2012.6398319
U. R. Aktas, G. Can, F. Vural
Shoreline extraction algorithms from multispectral imagery depend on threshold selection over spectral values and segmentation in general. Although this method gives high performance values for water delineation, error is accumulated on pixels near shoreline and complicates detection of nearby ships, docks etc. Water-shadow spectral mixing and spectral difference in water regions are two of the reasons for such untrustworthy shoreline results. With only four bands available, improvement in water detection depending only on pixel values is not very promising. Therefore, segmentation gains importance. By an edge-aware segmentation method, we aim to improve overall water and shoreline detection performances. In this study, a robust three-stage shoreline extraction algorithm is proposed. In the first stage, segmentation is applied over spectral values and then, some segments are combined according to edge information. In the second stage of the algorithm, pixel-based water information is combined with segmentation. The last step consists of enhancement of water regions based on local optimization by merging regions near shore boundary. Additionally, two new boundary-sensitive performance metrics are introduced for measuring the accuracy of the detected boundaries.
{"title":"Edge-aware segmentation in satellite imagery: A case study of shoreline detection","authors":"U. R. Aktas, G. Can, F. Vural","doi":"10.1109/PPRS.2012.6398319","DOIUrl":"https://doi.org/10.1109/PPRS.2012.6398319","url":null,"abstract":"Shoreline extraction algorithms from multispectral imagery depend on threshold selection over spectral values and segmentation in general. Although this method gives high performance values for water delineation, error is accumulated on pixels near shoreline and complicates detection of nearby ships, docks etc. Water-shadow spectral mixing and spectral difference in water regions are two of the reasons for such untrustworthy shoreline results. With only four bands available, improvement in water detection depending only on pixel values is not very promising. Therefore, segmentation gains importance. By an edge-aware segmentation method, we aim to improve overall water and shoreline detection performances. In this study, a robust three-stage shoreline extraction algorithm is proposed. In the first stage, segmentation is applied over spectral values and then, some segments are combined according to edge information. In the second stage of the algorithm, pixel-based water information is combined with segmentation. The last step consists of enhancement of water regions based on local optimization by merging regions near shore boundary. Additionally, two new boundary-sensitive performance metrics are introduced for measuring the accuracy of the detected boundaries.","PeriodicalId":139043,"journal":{"name":"7th IAPR Workshop on Pattern Recognition in Remote Sensing (PRRS)","volume":"09 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127287752","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2012-12-31DOI: 10.1109/PPRS.2012.6398321
E. Michaelsen
Gestalt-laws such as good continuation and similarity are coded as production systems. Applied to aerial images such systems can automatically perform grouping inferences following the archetype of human perception. Two variants are investigated on the same data: (1) Using only the geometrical gestalt laws on short contour line objects; (2) using also color close to the contours from the NIR images. The second variant tends to produce less illusory gestalts and more useful output.
{"title":"Perceptual grouping of row-gestalts in aerial NIR images of urban terrain","authors":"E. Michaelsen","doi":"10.1109/PPRS.2012.6398321","DOIUrl":"https://doi.org/10.1109/PPRS.2012.6398321","url":null,"abstract":"Gestalt-laws such as good continuation and similarity are coded as production systems. Applied to aerial images such systems can automatically perform grouping inferences following the archetype of human perception. Two variants are investigated on the same data: (1) Using only the geometrical gestalt laws on short contour line objects; (2) using also color close to the contours from the NIR images. The second variant tends to produce less illusory gestalts and more useful output.","PeriodicalId":139043,"journal":{"name":"7th IAPR Workshop on Pattern Recognition in Remote Sensing (PRRS)","volume":"226 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130971072","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2012-12-31DOI: 10.1109/PPRS.2012.6398322
Qianwen Yang, F. Sun, Huaping Liu
This article focuses on remote sensing image fusion in order to improve target recognition performance. Current fusion algorithms are mostly designed for specific purpose and have exponential complexity. We propose a fast and robust image fusion algorithm-the iterative learning fusion (ILF) algorithm, to improve the quality of images. This algorithm combines iterative learning in control theory with Multi-scale Geometric Analysis (MGA) image fusion algorithms; also, we apply color transfer to preserve color feature and cooperate it with SVM to improve recognition. By performing iterative learning, fusion parameters will converge to optimal in MGA fusion process. Theoretical analysis and experiments demonstrate improvement of visual and quantitative performance by proposed algorithm.
{"title":"Enhanced recognition using iterative learning fusion in remote sensing images","authors":"Qianwen Yang, F. Sun, Huaping Liu","doi":"10.1109/PPRS.2012.6398322","DOIUrl":"https://doi.org/10.1109/PPRS.2012.6398322","url":null,"abstract":"This article focuses on remote sensing image fusion in order to improve target recognition performance. Current fusion algorithms are mostly designed for specific purpose and have exponential complexity. We propose a fast and robust image fusion algorithm-the iterative learning fusion (ILF) algorithm, to improve the quality of images. This algorithm combines iterative learning in control theory with Multi-scale Geometric Analysis (MGA) image fusion algorithms; also, we apply color transfer to preserve color feature and cooperate it with SVM to improve recognition. By performing iterative learning, fusion parameters will converge to optimal in MGA fusion process. Theoretical analysis and experiments demonstrate improvement of visual and quantitative performance by proposed algorithm.","PeriodicalId":139043,"journal":{"name":"7th IAPR Workshop on Pattern Recognition in Remote Sensing (PRRS)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131606611","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2012-12-31DOI: 10.1109/PPRS.2012.6398316
H. Su, Peijun Du, Q. Du
Band clustering is applied to dimensionality reduction of hyperspectral imagery. The proposed method is based on a hierarchical clustering structure, which aims to group bands using an information or similarity measure. Specifically, the distance based on orthogonal projection divergence (OPD) is used as a criterion for clustering. Moreover, different from unsupervised clustering using all the pixels or supervised clustering requiring labeled pixels, the proposed semi-supervised band clustering needs class spectral signatures only. The experimental results show that the proposed algorithm can significantly outperform other existing methods with regard to pixel-based classification task.
{"title":"Hierarchical band clustering for hyperspectral image analysis","authors":"H. Su, Peijun Du, Q. Du","doi":"10.1109/PPRS.2012.6398316","DOIUrl":"https://doi.org/10.1109/PPRS.2012.6398316","url":null,"abstract":"Band clustering is applied to dimensionality reduction of hyperspectral imagery. The proposed method is based on a hierarchical clustering structure, which aims to group bands using an information or similarity measure. Specifically, the distance based on orthogonal projection divergence (OPD) is used as a criterion for clustering. Moreover, different from unsupervised clustering using all the pixels or supervised clustering requiring labeled pixels, the proposed semi-supervised band clustering needs class spectral signatures only. The experimental results show that the proposed algorithm can significantly outperform other existing methods with regard to pixel-based classification task.","PeriodicalId":139043,"journal":{"name":"7th IAPR Workshop on Pattern Recognition in Remote Sensing (PRRS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128011348","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2012-12-31DOI: 10.1109/PPRS.2012.6398311
L. Schack, A. Schunert, U. Soergel
Modern spaceborne SAR sensors like TerraSAR-X offer ground resolutions of about one meter in range and azimuth direction which allows for the discrimination between different facade elements. Those objects often feature a trihedral structure, which leads to a strong radar response due to triple-bounce reflection. The resulting radar signal is frequently observed to be temporally stable and usually shows a large amplitude (Persistent Scatterer, PS). Our aim is to aggregate single PS to lattices using Gestalt theory. This is a very important step in the process of unveiling the physical nature of PS since it simplifies the fusion with supplementary data like oblique view aerial images. We use a two stage 2D production system which exploits the knowledge abut mapping of 3D objects into the SAR imaging geometry. In the first step, the production system groups those PS, which are vertically aligned in the real world, to rows. In a second step, those groups are merged to regular lattices with the second orientation corresponding to the horizontal alignment of facade elements. The results show the benefit of aggregating points to lattices by the possible distinction of facade orientations.
{"title":"A two-dimensional production system for grouping persistent scatterers in urban high-resolution SAR scenes","authors":"L. Schack, A. Schunert, U. Soergel","doi":"10.1109/PPRS.2012.6398311","DOIUrl":"https://doi.org/10.1109/PPRS.2012.6398311","url":null,"abstract":"Modern spaceborne SAR sensors like TerraSAR-X offer ground resolutions of about one meter in range and azimuth direction which allows for the discrimination between different facade elements. Those objects often feature a trihedral structure, which leads to a strong radar response due to triple-bounce reflection. The resulting radar signal is frequently observed to be temporally stable and usually shows a large amplitude (Persistent Scatterer, PS). Our aim is to aggregate single PS to lattices using Gestalt theory. This is a very important step in the process of unveiling the physical nature of PS since it simplifies the fusion with supplementary data like oblique view aerial images. We use a two stage 2D production system which exploits the knowledge abut mapping of 3D objects into the SAR imaging geometry. In the first step, the production system groups those PS, which are vertically aligned in the real world, to rows. In a second step, those groups are merged to regular lattices with the second orientation corresponding to the horizontal alignment of facade elements. The results show the benefit of aggregating points to lattices by the possible distinction of facade orientations.","PeriodicalId":139043,"journal":{"name":"7th IAPR Workshop on Pattern Recognition in Remote Sensing (PRRS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123667226","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2012-12-31DOI: 10.1109/PPRS.2012.6398312
S. Kosov, F. Rottensteiner, C. Heipke
We apply Conditional Random Fields for the classification of scenes containing crossroads, using a simple appearance-based model in combination with a probabilistic model of the co-occurrence of class labels at neighbouring image sites. We use multiple overlap aerial images to derive a digital surface model and a true orthophoto without dynamic objects such as cars. An evaluation on an urban data set of aerial images delivers promising results.
{"title":"3D classification of crossroads from multiple aerial images using conditional random fields","authors":"S. Kosov, F. Rottensteiner, C. Heipke","doi":"10.1109/PPRS.2012.6398312","DOIUrl":"https://doi.org/10.1109/PPRS.2012.6398312","url":null,"abstract":"We apply Conditional Random Fields for the classification of scenes containing crossroads, using a simple appearance-based model in combination with a probabilistic model of the co-occurrence of class labels at neighbouring image sites. We use multiple overlap aerial images to derive a digital surface model and a true orthophoto without dynamic objects such as cars. An evaluation on an urban data set of aerial images delivers promising results.","PeriodicalId":139043,"journal":{"name":"7th IAPR Workshop on Pattern Recognition in Remote Sensing (PRRS)","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113954662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2012-12-31DOI: 10.1109/PPRS.2012.6398314
H. Shimoda, K. Fukue
The objective of this study is to develop high accuracy land cover classification algorithm for Global scale by using multi-temporal MODIS land reflectance products. In this study, time-domain co-occurrence matrix was introduced as a classification feature which provides time-series signature of land covers. Further, the non-parametric minimum distance classifier was introduced for time-domain co-occurrence matrix, which performs multi-dimensional pattern matching for time-domain co-occurrence matrices of a classification target pixel and each classification classes. The global land cover classification experiments have been conducted by applying the proposed classification method using 46 multi-temporal(in one year) SR(Surface Reflectance 8-Day L3) and NBAR(Nadir BRDF-Adjusted Reflectance 16-Day L3) products, respectively. IGBP 17 land cover categories were used in our classification experiments. As the results, SR product and NBAR product showed similar classification accuracy of 99%.
{"title":"Global land cover classification using MODIS surface reflectance products","authors":"H. Shimoda, K. Fukue","doi":"10.1109/PPRS.2012.6398314","DOIUrl":"https://doi.org/10.1109/PPRS.2012.6398314","url":null,"abstract":"The objective of this study is to develop high accuracy land cover classification algorithm for Global scale by using multi-temporal MODIS land reflectance products. In this study, time-domain co-occurrence matrix was introduced as a classification feature which provides time-series signature of land covers. Further, the non-parametric minimum distance classifier was introduced for time-domain co-occurrence matrix, which performs multi-dimensional pattern matching for time-domain co-occurrence matrices of a classification target pixel and each classification classes. The global land cover classification experiments have been conducted by applying the proposed classification method using 46 multi-temporal(in one year) SR(Surface Reflectance 8-Day L3) and NBAR(Nadir BRDF-Adjusted Reflectance 16-Day L3) products, respectively. IGBP 17 land cover categories were used in our classification experiments. As the results, SR product and NBAR product showed similar classification accuracy of 99%.","PeriodicalId":139043,"journal":{"name":"7th IAPR Workshop on Pattern Recognition in Remote Sensing (PRRS)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130172203","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2012-12-31DOI: 10.1109/PPRS.2012.6398313
Y. Nakayama, Y. Hara, K. Endo
The purpose of this study is to analyze the distribution and the classification of gravel bar along the shore of Balkhash Lake by using the high precision DSM newly produced from the stereo pair data of PRISM carried in ALOS, and to consider the age and environment of formation of every classified gravel bar, and the long-term environmental change based on water level fluctuation of the lake. The classification of gravel bar distribution pattern based on the characteristics such as height, number and direction of the ridge line was carried out through extraction and comparative analysis of the DSM data and ALOS pan-sharpened image in eighteen areas along the shoreline. The result of this study showed that the distribution pattern of the gravel bars along the shore was classified into three typical groups. The difference in distribution of three groups is based on the influence of the waves with the strong prevailing wind. Moreover, according to the feature of the classified group, the water level change situation of Balkhash Lake was divided into three stages in the past about 26,000 years, and the tendency of drawdown was shown as the whole.
{"title":"Classification of microtopographical features along the shore of Balkhash Lake by ALOS/PRISM DSM","authors":"Y. Nakayama, Y. Hara, K. Endo","doi":"10.1109/PPRS.2012.6398313","DOIUrl":"https://doi.org/10.1109/PPRS.2012.6398313","url":null,"abstract":"The purpose of this study is to analyze the distribution and the classification of gravel bar along the shore of Balkhash Lake by using the high precision DSM newly produced from the stereo pair data of PRISM carried in ALOS, and to consider the age and environment of formation of every classified gravel bar, and the long-term environmental change based on water level fluctuation of the lake. The classification of gravel bar distribution pattern based on the characteristics such as height, number and direction of the ridge line was carried out through extraction and comparative analysis of the DSM data and ALOS pan-sharpened image in eighteen areas along the shoreline. The result of this study showed that the distribution pattern of the gravel bars along the shore was classified into three typical groups. The difference in distribution of three groups is based on the influence of the waves with the strong prevailing wind. Moreover, according to the feature of the classified group, the water level change situation of Balkhash Lake was divided into three stages in the past about 26,000 years, and the tendency of drawdown was shown as the whole.","PeriodicalId":139043,"journal":{"name":"7th IAPR Workshop on Pattern Recognition in Remote Sensing (PRRS)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131150649","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2012-12-31DOI: 10.1109/PPRS.2012.6398309
F. de Morsier, D. Tuia, V. Gass, J. Thiran, M. Borgeaud
When dealing with change detection problems, information about the nature of the changes is often unavailable. In this paper we propose a solution to perform unsupervised change detection based on nonlinear support vector clustering. We build a series of nested hierarchical support vector clustering descriptions, select the appropriate one using a cluster validity measure and finally merge the clusters into two classes, corresponding to changed and unchanged areas. Experiments on two multispectral datasets confirm the power and appropriateness of the proposed system.
{"title":"Unsupervised change detection via hierarchical support vector clustering","authors":"F. de Morsier, D. Tuia, V. Gass, J. Thiran, M. Borgeaud","doi":"10.1109/PPRS.2012.6398309","DOIUrl":"https://doi.org/10.1109/PPRS.2012.6398309","url":null,"abstract":"When dealing with change detection problems, information about the nature of the changes is often unavailable. In this paper we propose a solution to perform unsupervised change detection based on nonlinear support vector clustering. We build a series of nested hierarchical support vector clustering descriptions, select the appropriate one using a cluster validity measure and finally merge the clusters into two classes, corresponding to changed and unchanged areas. Experiments on two multispectral datasets confirm the power and appropriateness of the proposed system.","PeriodicalId":139043,"journal":{"name":"7th IAPR Workshop on Pattern Recognition in Remote Sensing (PRRS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115973627","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2012-12-31DOI: 10.1109/PPRS.2012.6398310
Xuezhi Yang, Li Jia, Yujie Wang, Yiming Tang
In this paper, a new metric called despeckling structural loss(DSL) is proposed for performance assessment of despeckling algorithms with a focus on the preservation of structural information. By taking into account characteristics of the best and worst structure preservation in despeckling, the DSL metric examines the presence of image structures in ratio images by using local correlations between the ratio image and the noise-free reference image at edge points, leading an objective and quantitative measure of the structure-preserving capability of despeckling algorithms. The DSL metric has been tested on despeckling results of a simulated SAR image using three types of algorithms and efficiency of the DSL has been demonstrated. In comparison, the other five commonly used despeckling metrics fail to keep a consistency with the structural loss shown in despeckling results as well as ratio images.
{"title":"Despeckling structural loss(DSL): A new metric for measuring structure-preserving capability of despeckling algorithms","authors":"Xuezhi Yang, Li Jia, Yujie Wang, Yiming Tang","doi":"10.1109/PPRS.2012.6398310","DOIUrl":"https://doi.org/10.1109/PPRS.2012.6398310","url":null,"abstract":"In this paper, a new metric called despeckling structural loss(DSL) is proposed for performance assessment of despeckling algorithms with a focus on the preservation of structural information. By taking into account characteristics of the best and worst structure preservation in despeckling, the DSL metric examines the presence of image structures in ratio images by using local correlations between the ratio image and the noise-free reference image at edge points, leading an objective and quantitative measure of the structure-preserving capability of despeckling algorithms. The DSL metric has been tested on despeckling results of a simulated SAR image using three types of algorithms and efficiency of the DSL has been demonstrated. In comparison, the other five commonly used despeckling metrics fail to keep a consistency with the structural loss shown in despeckling results as well as ratio images.","PeriodicalId":139043,"journal":{"name":"7th IAPR Workshop on Pattern Recognition in Remote Sensing (PRRS)","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133184746","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：