When travelling,people are accustomed to taking and uploading photos on social media websites,which has led to the accumulation of huge numbers of geotagged photos.Combined with multisource information(e.g.weather,tra...When travelling,people are accustomed to taking and uploading photos on social media websites,which has led to the accumulation of huge numbers of geotagged photos.Combined with multisource information(e.g.weather,transportation,or textual information),these geotagged photos could help us in constructing user preference profiles at a high level of detail.Therefore,using these geotagged photos,we built a personalised recommendation system to provide attraction recommendations that match a user’s preferences.Specifically,we retrieved a geotagged photo collection from the public API for Flickr(Flickr.com)and fetched a large amount of other contextual information to rebuild a user’s travel history.We then created a model-based recommendation method with a two-stage architecture that consists of candidate generation(the matching process)and candidate ranking.In the matching process,we used a support vector machine model that was modified for multiclass classification to generate the candidate list.In addition,we used a gradient boosting regression tree to score each candidate and rerank the list.Finally,we evaluated our recommendation results with respect to accuracy and ranking ability.Compared with widely used memory-based methods,our proposed method performs significantly better in the cold-start situation and when mining‘long-tail’data.展开更多
Forest fires in mountainous areas can cause severe defores-tation which can potentially trigger secondary natural hazards like debris falls and avalanches. We documented an extreme case study for the range of possible...Forest fires in mountainous areas can cause severe defores-tation which can potentially trigger secondary natural hazards like debris falls and avalanches. We documented an extreme case study for the range of possible post-fire land cover （LC） dynamics. We investigated a 15-ha, steep （10°-65°） burnt slope in Styria （Austria） at elevation of 760-1130 m, which burned in 1946 and has not fully recovered to date. Seven 8-class legend LC maps were produced （1954, 1966, 1973, 1982, 1998, 2004, 2009） and integrated in a vector-based GIS, mainly by on-screen interpretation of aerial photos. Our aim was to clarify how post-wildfire LC dynamics take place on a severely damaged, steep slope and to give a basic projection of the future vegetation recovery process. The pre-fire Pinus sylvestrisstands have been mainly replaced by Picea abies and Larix decidua. Regeneration proceeded mainly from the base of the slope upwards. All tree species together still cover no more than 40% of the slope after more than 60 years of recovery, while grassland communities and rock/debris areas have expanded. Multitemporal analysis showed a slow but steady increase in woodland cover. Degraded rock/debris areas, however, expanded as well because soil erosion and related debris flows remained active. Slope angle （with a threshold value of approx. 35-40°） seemed to control whether erosion or regeneration prevailed. According to a simple extrapolation, the slope will not reach its former condition before 2070. This extreme disturbance window of more than 120 years is owed to the steepness of the slope and to the shallow soils on dolomitic bedrock that were severely damaged by the fire. The neglect of any game fencing is a further factor slowing regeneration.展开更多
The objective of style transfer is to maintain the content of an image while transferring the style of another image.However,conventional methods face challenges in preserving facial features,especially in Korean port...The objective of style transfer is to maintain the content of an image while transferring the style of another image.However,conventional methods face challenges in preserving facial features,especially in Korean portraits where elements like the“Gat”(a traditional Korean hat)are prevalent.This paper proposes a deep learning network designed to perform style transfer that includes the“Gat”while preserving the identity of the face.Unlike traditional style transfer techniques,the proposed method aims to preserve the texture,attire,and the“Gat”in the style image by employing image sharpening and face landmark,with the GAN.The color,texture,and intensity were extracted differently based on the characteristics of each block and layer of the pre-trained VGG-16,and only the necessary elements during training were preserved using a facial landmark mask.The head area was presented using the eyebrow area to transfer the“Gat”.Furthermore,the identity of the face was retained,and style correlation was considered based on the Gram matrix.To evaluate performance,we introduced a metric using PSNR and SSIM,with an emphasis on median values through new weightings for style transfer in Korean portraits.Additionally,we have conducted a survey that evaluated the content,style,and naturalness of the transferred results,and based on the assessment,we can confidently conclude that our method to maintain the integrity of content surpasses the previous research.Our approach,enriched by landmarks preservation and diverse loss functions,including those related to“Gat”,outperformed previous researches in facial identity preservation.展开更多
The aim of our study was to assess the influence of motorways on landscape structure changes. The research was situated on segments of the D1 and D3 motorways in the Czech Republic. The method used for determining the...The aim of our study was to assess the influence of motorways on landscape structure changes. The research was situated on segments of the D1 and D3 motorways in the Czech Republic. The method used for determining the impacts of the construction and operation of the motorway on the landscape structure was based on comparing the development of the land cover types in an area where the motorway was present for almost the entire monitored period (D1 motorway segment) with the development of the land cover types in an area where the motorway was constructed towards the end of the monitored period (D3 motorway segment). Monitored period?was?between years 1949, 1988 (1984 segment of motorway D3) and 2007. Source materials were processed and analyzed in the ArcGIS software environment. Our comparison indicated that the main differences between segments were in the level of development of commercial, industrial and agricultural built-up area in the vicinity of motorways. The existence of the motorway also contributed to the development of residential areas. It should be noted that in both cases there was also an increase in the extension?of shrubs and scattered vegetation, and in small quantities also category of forest.展开更多
Cross-regional locality research reflects the influences of natural environment and the human activities due to the abundant land types and the multiple landscape combinations in related regions. The Chinese farming-p...Cross-regional locality research reflects the influences of natural environment and the human activities due to the abundant land types and the multiple landscape combinations in related regions. The Chinese farming-pastoral ecotone is a typical large-scale region but few studies were conducted. This research contributed to the understanding of cross-regional locality of the Chinese farming-pastoral ecotone from different scales, including national, sectional, and provincial administrative units by utilizing geotagged photos（GTPs） obtained from the Panoramio website. The major results were as follows：（1） the locality elements of the Chinese farming-pastoral ecotone included 52 free nodes classified into 8 types of scene attributes;（2） there were huge differences between locality elements of different regions, and there was a negative correlation between the similarity degree of elements of different provinces and their spatial distances;（3） the Chinese farming-pastoral ecotone could be divided into the northern, central and southern sections, whose localities had differences in element constitution, association structure and the strength of elements, system stability and the anti-interference capability; and（4） the evolution of the localities of the northern and central sections was mainly influenced by human activities, while the locality of southern section retained more natural features. On a theoretical level, this research aimed to establish the research methodology of locality from the perspective of open data on the web with strong operability and replicability. On a practical level, this research could enrich the structuring recognition of the locality of the Chinese farming-pastoral ecotone and the comprehension of its dynamic mechanism. The results provide a reference for locality differentiation protection and the development of a cross-regional scale.展开更多
The generation of reality-based 3D models of archeological sites is the future of representing existing ancient structures. Such approach requires substantial economic and logistical costs which limit this application...The generation of reality-based 3D models of archeological sites is the future of representing existing ancient structures. Such approach requires substantial economic and logistical costs which limit this application. In this context, this paper presents the use of photogrammetric workflow, based on Structure from Motion techniques (SfM) to derive 3D metric information from Structure-and-Motion images. The interdisciplinary 3D modeling framework consists of fusion of multi resolution images from both overlapped low-altitude aerial and multi-view terrestrial imagery. The acquisition of aerial photographs survey was based on archived oblique aerial stereo pair photos acquired from the Aerial Photographic Archive for Archaeology in the Middle East (APAAME) project, while terrestrial of close range photos covering the castle walls has been acquired using portable camera. Camera calibration and orientation were carried out by VisualSFM, CMPMVS (Multi-View Reconstruction) and SURE (Photogrammetric Surface Reconstruction from Imagery) software. The resulted cloud points were processed using cloud Compare, MeshLab, Agisoft, and Skethup Softwares. A complete 3D digital geometric recording of the site was accomplished based on dense 3D point clouds with realistic metric accuracy and photorealistic performance to meet all the surveying and archaeological needs. The final Ajloun castle’s digital model geometry was added as a 3D building layer on Google Earth.展开更多
Rapid urbanization and population growth of the Amman Area were combined with land resource degradation when the city was heading for mounting urbanization from the early 1950s. The deterioration of natural resources ...Rapid urbanization and population growth of the Amman Area were combined with land resource degradation when the city was heading for mounting urbanization from the early 1950s. The deterioration of natural resources and green areas, was coupled with the creation of different urban public open spaces in the city. The transformation from large single-family houses to dense residential apartments was associated with social and behavioral changes among the inhabitants living in the dense apartments. Consequently, a large private sector has been developed to provide public and social spaces. Photo-interpretation and satellite images were used to map and characterize land use/cover changes have been occurred in the Amman area between 1953 and 2017. Maximum Likelihood Classification method was employed to identify land use/cover changes between 1986 and 20017, and GIS was utilized also to map examples of the recently emerged socio-economic open spaces. Excessive urban development in the last two decades, and the adoption of neo-liberal privatization policies by the government, enhanced social stratification and residential segregation. So, instead of encouraging outdoor activity and social interaction among all groups of Amman residents, the freedom of accessibility to major open spaces has been restricted within the same urban fabric, thus, the “two Ammans” paradigm, a “tale of two cities”, has been recently acknowledged.展开更多
The present research study assesses the accuracy of two recent freely available ASTER-GDEM ver.2, SRTM ver.4.1 digital elevation models for parts of the lowest area on the earth on the shores of the Dead Sea, the site...The present research study assesses the accuracy of two recent freely available ASTER-GDEM ver.2, SRTM ver.4.1 digital elevation models for parts of the lowest area on the earth on the shores of the Dead Sea, the site located over areas with high variation in topography within the “humratessahan” watershed, where the elevation varies between 700 m above sea level in the north east of the watershed to 210 meter below sea level. The reference elevation data used in this study are topographical points extracted from existing digital vector topographical map published by Royal Jordanian Geographical Center (RJGC). However, the contour lines do not cover the whole area especially in the flat areas where elevation is less than the contour line interval. Therefore, the second reference data created with photogrammetric techniques from stereoscopic pairs of aerial photos had been used for the surface to surface comparison and elevation profiles assessment purpose. It is demonstrated that the vertical accuracy of ASTER-GDEM ver.2 is 22.223 m (Root Mean Square Error (RMSE)) against reference elevation data, while the SRTM ver.4.1 has an RMSE of 15.858 m;the statistics results indicate that the theoretical accuracy is achievable and meets the expected accuracy specification. Furthermore, as for its unprecedented detail, it is believed that the photogrammetric derived DEM offers an alternative in accessibility to high-quality elevation data with vertical accuracy of 0.78 m (Root Mean Square Error (RMSE)).展开更多
Becare of their low transmission loss, large bandwidth and small size, silica optical fibres have revolutionised telecommunications. Applications in other fields have been more limited. However, they are being increas...Becare of their low transmission loss, large bandwidth and small size, silica optical fibres have revolutionised telecommunications. Applications in other fields have been more limited. However, they are being increasingly used in military systems where their additional advantages of light weight, ruggedness, flexibility high strength and freedom electromagnetic interference can be erectal. The paper discusses a number of existing and proposed applications.展开更多
基金supported by grants from the National Key Research and Development Program of China[grant number 2017YFB0503602]the National Natural Science Foundation of China[grant number 41771425],[grant number 41625003],[grant number 41501162]the Beijing Philosophy and Social Science Foundation[grant number 17JDGLB002].
文摘When travelling,people are accustomed to taking and uploading photos on social media websites,which has led to the accumulation of huge numbers of geotagged photos.Combined with multisource information(e.g.weather,transportation,or textual information),these geotagged photos could help us in constructing user preference profiles at a high level of detail.Therefore,using these geotagged photos,we built a personalised recommendation system to provide attraction recommendations that match a user’s preferences.Specifically,we retrieved a geotagged photo collection from the public API for Flickr(Flickr.com)and fetched a large amount of other contextual information to rebuild a user’s travel history.We then created a model-based recommendation method with a two-stage architecture that consists of candidate generation(the matching process)and candidate ranking.In the matching process,we used a support vector machine model that was modified for multiclass classification to generate the candidate list.In addition,we used a gradient boosting regression tree to score each candidate and rerank the list.Finally,we evaluated our recommendation results with respect to accuracy and ranking ability.Compared with widely used memory-based methods,our proposed method performs significantly better in the cold-start situation and when mining‘long-tail’data.
文摘Forest fires in mountainous areas can cause severe defores-tation which can potentially trigger secondary natural hazards like debris falls and avalanches. We documented an extreme case study for the range of possible post-fire land cover （LC） dynamics. We investigated a 15-ha, steep （10°-65°） burnt slope in Styria （Austria） at elevation of 760-1130 m, which burned in 1946 and has not fully recovered to date. Seven 8-class legend LC maps were produced （1954, 1966, 1973, 1982, 1998, 2004, 2009） and integrated in a vector-based GIS, mainly by on-screen interpretation of aerial photos. Our aim was to clarify how post-wildfire LC dynamics take place on a severely damaged, steep slope and to give a basic projection of the future vegetation recovery process. The pre-fire Pinus sylvestrisstands have been mainly replaced by Picea abies and Larix decidua. Regeneration proceeded mainly from the base of the slope upwards. All tree species together still cover no more than 40% of the slope after more than 60 years of recovery, while grassland communities and rock/debris areas have expanded. Multitemporal analysis showed a slow but steady increase in woodland cover. Degraded rock/debris areas, however, expanded as well because soil erosion and related debris flows remained active. Slope angle （with a threshold value of approx. 35-40°） seemed to control whether erosion or regeneration prevailed. According to a simple extrapolation, the slope will not reach its former condition before 2070. This extreme disturbance window of more than 120 years is owed to the steepness of the slope and to the shallow soils on dolomitic bedrock that were severely damaged by the fire. The neglect of any game fencing is a further factor slowing regeneration.
基金supported by Metaverse Lab Program funded by the Ministry of Science and ICT(MSIT),and the Korea Radio Promotion Association(RAPA).
文摘The objective of style transfer is to maintain the content of an image while transferring the style of another image.However,conventional methods face challenges in preserving facial features,especially in Korean portraits where elements like the“Gat”(a traditional Korean hat)are prevalent.This paper proposes a deep learning network designed to perform style transfer that includes the“Gat”while preserving the identity of the face.Unlike traditional style transfer techniques,the proposed method aims to preserve the texture,attire,and the“Gat”in the style image by employing image sharpening and face landmark,with the GAN.The color,texture,and intensity were extracted differently based on the characteristics of each block and layer of the pre-trained VGG-16,and only the necessary elements during training were preserved using a facial landmark mask.The head area was presented using the eyebrow area to transfer the“Gat”.Furthermore,the identity of the face was retained,and style correlation was considered based on the Gram matrix.To evaluate performance,we introduced a metric using PSNR and SSIM,with an emphasis on median values through new weightings for style transfer in Korean portraits.Additionally,we have conducted a survey that evaluated the content,style,and naturalness of the transferred results,and based on the assessment,we can confidently conclude that our method to maintain the integrity of content surpasses the previous research.Our approach,enriched by landmarks preservation and diverse loss functions,including those related to“Gat”,outperformed previous researches in facial identity preservation.
文摘The aim of our study was to assess the influence of motorways on landscape structure changes. The research was situated on segments of the D1 and D3 motorways in the Czech Republic. The method used for determining the impacts of the construction and operation of the motorway on the landscape structure was based on comparing the development of the land cover types in an area where the motorway was present for almost the entire monitored period (D1 motorway segment) with the development of the land cover types in an area where the motorway was constructed towards the end of the monitored period (D3 motorway segment). Monitored period?was?between years 1949, 1988 (1984 segment of motorway D3) and 2007. Source materials were processed and analyzed in the ArcGIS software environment. Our comparison indicated that the main differences between segments were in the level of development of commercial, industrial and agricultural built-up area in the vicinity of motorways. The existence of the motorway also contributed to the development of residential areas. It should be noted that in both cases there was also an increase in the extension?of shrubs and scattered vegetation, and in small quantities also category of forest.
基金supported by the Sino-German Center (the National Natural Science Foundation of China and the German Science Foundation GZ1201)the Postgraduate Courses Project of Peking University (2014-40)
文摘Cross-regional locality research reflects the influences of natural environment and the human activities due to the abundant land types and the multiple landscape combinations in related regions. The Chinese farming-pastoral ecotone is a typical large-scale region but few studies were conducted. This research contributed to the understanding of cross-regional locality of the Chinese farming-pastoral ecotone from different scales, including national, sectional, and provincial administrative units by utilizing geotagged photos（GTPs） obtained from the Panoramio website. The major results were as follows：（1） the locality elements of the Chinese farming-pastoral ecotone included 52 free nodes classified into 8 types of scene attributes;（2） there were huge differences between locality elements of different regions, and there was a negative correlation between the similarity degree of elements of different provinces and their spatial distances;（3） the Chinese farming-pastoral ecotone could be divided into the northern, central and southern sections, whose localities had differences in element constitution, association structure and the strength of elements, system stability and the anti-interference capability; and（4） the evolution of the localities of the northern and central sections was mainly influenced by human activities, while the locality of southern section retained more natural features. On a theoretical level, this research aimed to establish the research methodology of locality from the perspective of open data on the web with strong operability and replicability. On a practical level, this research could enrich the structuring recognition of the locality of the Chinese farming-pastoral ecotone and the comprehension of its dynamic mechanism. The results provide a reference for locality differentiation protection and the development of a cross-regional scale.
文摘The generation of reality-based 3D models of archeological sites is the future of representing existing ancient structures. Such approach requires substantial economic and logistical costs which limit this application. In this context, this paper presents the use of photogrammetric workflow, based on Structure from Motion techniques (SfM) to derive 3D metric information from Structure-and-Motion images. The interdisciplinary 3D modeling framework consists of fusion of multi resolution images from both overlapped low-altitude aerial and multi-view terrestrial imagery. The acquisition of aerial photographs survey was based on archived oblique aerial stereo pair photos acquired from the Aerial Photographic Archive for Archaeology in the Middle East (APAAME) project, while terrestrial of close range photos covering the castle walls has been acquired using portable camera. Camera calibration and orientation were carried out by VisualSFM, CMPMVS (Multi-View Reconstruction) and SURE (Photogrammetric Surface Reconstruction from Imagery) software. The resulted cloud points were processed using cloud Compare, MeshLab, Agisoft, and Skethup Softwares. A complete 3D digital geometric recording of the site was accomplished based on dense 3D point clouds with realistic metric accuracy and photorealistic performance to meet all the surveying and archaeological needs. The final Ajloun castle’s digital model geometry was added as a 3D building layer on Google Earth.
文摘Rapid urbanization and population growth of the Amman Area were combined with land resource degradation when the city was heading for mounting urbanization from the early 1950s. The deterioration of natural resources and green areas, was coupled with the creation of different urban public open spaces in the city. The transformation from large single-family houses to dense residential apartments was associated with social and behavioral changes among the inhabitants living in the dense apartments. Consequently, a large private sector has been developed to provide public and social spaces. Photo-interpretation and satellite images were used to map and characterize land use/cover changes have been occurred in the Amman area between 1953 and 2017. Maximum Likelihood Classification method was employed to identify land use/cover changes between 1986 and 20017, and GIS was utilized also to map examples of the recently emerged socio-economic open spaces. Excessive urban development in the last two decades, and the adoption of neo-liberal privatization policies by the government, enhanced social stratification and residential segregation. So, instead of encouraging outdoor activity and social interaction among all groups of Amman residents, the freedom of accessibility to major open spaces has been restricted within the same urban fabric, thus, the “two Ammans” paradigm, a “tale of two cities”, has been recently acknowledged.
文摘The present research study assesses the accuracy of two recent freely available ASTER-GDEM ver.2, SRTM ver.4.1 digital elevation models for parts of the lowest area on the earth on the shores of the Dead Sea, the site located over areas with high variation in topography within the “humratessahan” watershed, where the elevation varies between 700 m above sea level in the north east of the watershed to 210 meter below sea level. The reference elevation data used in this study are topographical points extracted from existing digital vector topographical map published by Royal Jordanian Geographical Center (RJGC). However, the contour lines do not cover the whole area especially in the flat areas where elevation is less than the contour line interval. Therefore, the second reference data created with photogrammetric techniques from stereoscopic pairs of aerial photos had been used for the surface to surface comparison and elevation profiles assessment purpose. It is demonstrated that the vertical accuracy of ASTER-GDEM ver.2 is 22.223 m (Root Mean Square Error (RMSE)) against reference elevation data, while the SRTM ver.4.1 has an RMSE of 15.858 m;the statistics results indicate that the theoretical accuracy is achievable and meets the expected accuracy specification. Furthermore, as for its unprecedented detail, it is believed that the photogrammetric derived DEM offers an alternative in accessibility to high-quality elevation data with vertical accuracy of 0.78 m (Root Mean Square Error (RMSE)).
文摘Becare of their low transmission loss, large bandwidth and small size, silica optical fibres have revolutionised telecommunications. Applications in other fields have been more limited. However, they are being increasingly used in military systems where their additional advantages of light weight, ruggedness, flexibility high strength and freedom electromagnetic interference can be erectal. The paper discusses a number of existing and proposed applications.
