Pub Date : 2024-05-11DOI: 10.5194/isprs-archives-xlviii-1-2024-927-2024
Jong-Hwan Son, Sumin Park, Hyeon-Ju Ban, Taejung Kim
Abstract. High-resolution satellite imagery has a limitation in terms of coverage area. This limitation presents challenges for extensive-scale analysis at regional or national levels. To maximize the utility of high-resolution satellite imagery, the implementation of image mosaicking techniques is essential. In this paper, we have developed seamline extraction techniques and relative geometric correction optimized for high-resolution satellite imagery. Ultimately, we proposed a multi-strip image mosaicking method for KOMPSAT-3A (Korea Multi-Purpose Satellite-3A) images. We applied the Dijkstra's shortest path algorithm to efficiently extract seamlines. we also performed image registration based on feature matching and homography transformation to correct the relative geometric errors between input images. We conducted experiments with our methods using 29 scenes from KOMPSAT-3A L1G data. The results indicated high relative geometric accuracy, with an average error of 1.63 pixels. Furthermore, we were able to obtain high-quality seamless mosaic images. Our proposed method is expected to enhance the utility of KOMPSAT-3A imagery for large-scale environmental and urban analysis and to provide more accurate and comprehensive data.
{"title":"Development of Multi-Strip Image Mosaicking for KOMPSAT-3A Images","authors":"Jong-Hwan Son, Sumin Park, Hyeon-Ju Ban, Taejung Kim","doi":"10.5194/isprs-archives-xlviii-1-2024-927-2024","DOIUrl":"https://doi.org/10.5194/isprs-archives-xlviii-1-2024-927-2024","url":null,"abstract":"Abstract. High-resolution satellite imagery has a limitation in terms of coverage area. This limitation presents challenges for extensive-scale analysis at regional or national levels. To maximize the utility of high-resolution satellite imagery, the implementation of image mosaicking techniques is essential. In this paper, we have developed seamline extraction techniques and relative geometric correction optimized for high-resolution satellite imagery. Ultimately, we proposed a multi-strip image mosaicking method for KOMPSAT-3A (Korea Multi-Purpose Satellite-3A) images. We applied the Dijkstra's shortest path algorithm to efficiently extract seamlines. we also performed image registration based on feature matching and homography transformation to correct the relative geometric errors between input images. We conducted experiments with our methods using 29 scenes from KOMPSAT-3A L1G data. The results indicated high relative geometric accuracy, with an average error of 1.63 pixels. Furthermore, we were able to obtain high-quality seamless mosaic images. Our proposed method is expected to enhance the utility of KOMPSAT-3A imagery for large-scale environmental and urban analysis and to provide more accurate and comprehensive data.\u0000","PeriodicalId":505918,"journal":{"name":"The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences","volume":" 41","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140990433","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 : 2024-05-11DOI: 10.5194/isprs-archives-xlviii-1-2024-843-2024
Jinian Zhang, Lanfa Liu
Abstract. Simulating land use and land cover changes (LUCC) is important for urban planning and environmental studies. In this study, we introduce a neural cellular automata (NCA) model that integrates biological principles and convolutional neural networks (CNNs) for land use simulation. We conduct experiments in the city of Wuhan, China. The NCA model achieved the highest performance with an OA of 0.858, F1 score of 0.753, Kappa coefficient of 0.799, and FOM of 0.427. Comparisons of land use data of Wuhan city from 2000 and 2010 with the simulated optimal results indicate that forest areas closer to urban centers are more susceptible to modernization processes, showing the advantage of NCA in accurately simulating land use changes in the central urban area.
{"title":"Neural Cellular Automata-based Land Use Changes Simulation","authors":"Jinian Zhang, Lanfa Liu","doi":"10.5194/isprs-archives-xlviii-1-2024-843-2024","DOIUrl":"https://doi.org/10.5194/isprs-archives-xlviii-1-2024-843-2024","url":null,"abstract":"Abstract. Simulating land use and land cover changes (LUCC) is important for urban planning and environmental studies. In this study, we introduce a neural cellular automata (NCA) model that integrates biological principles and convolutional neural networks (CNNs) for land use simulation. We conduct experiments in the city of Wuhan, China. The NCA model achieved the highest performance with an OA of 0.858, F1 score of 0.753, Kappa coefficient of 0.799, and FOM of 0.427. Comparisons of land use data of Wuhan city from 2000 and 2010 with the simulated optimal results indicate that forest areas closer to urban centers are more susceptible to modernization processes, showing the advantage of NCA in accurately simulating land use changes in the central urban area.\u0000","PeriodicalId":505918,"journal":{"name":"The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences","volume":" 1152","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140988732","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 : 2024-05-11DOI: 10.5194/isprs-archives-xlviii-1-2024-697-2024
Xiaying Wang, Shuaiqiang Chen, Yuanping Xia, Yufen Niu, Jun Gong, Yumei Yang
Abstract. Ground surface deformation in mines affects mining production, damages the ecological environment, and endangers human safety. Mastering the detailed time series deformation and related triggering factors can provide key information for the safety of the mining area. Therefore, the Fengcheng mining area, a large and ancient coal mine in Jiangxi Province, China, was selected as the study area in this work, and the following research was conducted: 1. The accuracy and applicability of the Stacking, Small-Baseline Subset InSAR (SBAS-InSAR), and Interferometric Point Target Analysis (IPTA) methods were preliminarily explored while monitoring the annual deformation rate based on Sentinel-1A data from October 2019 to November 2022. 2. The time-series deformation of the Fengcheng mining area was obtained with SBAS-InSAR technology, and the sedimentation was validated with leveling results. 3. The correlation factors of deformation, such as rainfall and land cover, were studied, and the relationship between the influencing factors, such as coal mining dip angle, digital elevation, coal mining elevation, and deformation, was quantitatively explored with the Grey correlation model and Pearson correlation analysis method. The following conclusions were drawn: The SBAS method has the best adaptability in the dense vegetation mining area, and the root-mean-square error of the difference between deformation results and leveling data does not exceed 4mm. The evolution process of deformation centers is mainly divided into the stages of initial deformation, constant velocity deformation, accelerated deformation, and stable condition. Compared with the natural factors, the settlement of the Fengcheng mining area is mainly affected by human-induced mining and construction of artificial facilities.
{"title":"Analysis of surface deformation and related factors over mining areas based on InSAR: A case study of Fengcheng mine","authors":"Xiaying Wang, Shuaiqiang Chen, Yuanping Xia, Yufen Niu, Jun Gong, Yumei Yang","doi":"10.5194/isprs-archives-xlviii-1-2024-697-2024","DOIUrl":"https://doi.org/10.5194/isprs-archives-xlviii-1-2024-697-2024","url":null,"abstract":"Abstract. Ground surface deformation in mines affects mining production, damages the ecological environment, and endangers human safety. Mastering the detailed time series deformation and related triggering factors can provide key information for the safety of the mining area. Therefore, the Fengcheng mining area, a large and ancient coal mine in Jiangxi Province, China, was selected as the study area in this work, and the following research was conducted: 1. The accuracy and applicability of the Stacking, Small-Baseline Subset InSAR (SBAS-InSAR), and Interferometric Point Target Analysis (IPTA) methods were preliminarily explored while monitoring the annual deformation rate based on Sentinel-1A data from October 2019 to November 2022. 2. The time-series deformation of the Fengcheng mining area was obtained with SBAS-InSAR technology, and the sedimentation was validated with leveling results. 3. The correlation factors of deformation, such as rainfall and land cover, were studied, and the relationship between the influencing factors, such as coal mining dip angle, digital elevation, coal mining elevation, and deformation, was quantitatively explored with the Grey correlation model and Pearson correlation analysis method. The following conclusions were drawn: The SBAS method has the best adaptability in the dense vegetation mining area, and the root-mean-square error of the difference between deformation results and leveling data does not exceed 4mm. The evolution process of deformation centers is mainly divided into the stages of initial deformation, constant velocity deformation, accelerated deformation, and stable condition. Compared with the natural factors, the settlement of the Fengcheng mining area is mainly affected by human-induced mining and construction of artificial facilities.\u0000","PeriodicalId":505918,"journal":{"name":"The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences","volume":" 11","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140988799","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 : 2024-05-10DOI: 10.5194/isprs-archives-xlviii-1-2024-129-2024
Shuai Dong, Chenni Lu, Wenchao Gao, Chang Liu, Jin Bai
Abstract. In order to improve the data quality of land use remote sensing monitoring images, this article introduces the process of generating satellite image data, elaborates on the content of satellite image data verification in land use remote sensing monitoring, and proposes quality issues and improvement measures for satellite image data. Taking the discovered satellite image data quality issues as an example, compared with quality inspection standards, it was found that the main problems in the results were projection parameter errors, image color distortion, image blurring, and position accuracy exceeding limits. It is recommended to check the above issues during the image production stage, analyze the reasons for exceeding the position accuracy limit, image distortion, and embossing, and provide relevant suggestions. Provided strong technical support for land use surveys.
{"title":"Quality Inspection and Problem Analysis of Satellite Image Data in Land Use Survey","authors":"Shuai Dong, Chenni Lu, Wenchao Gao, Chang Liu, Jin Bai","doi":"10.5194/isprs-archives-xlviii-1-2024-129-2024","DOIUrl":"https://doi.org/10.5194/isprs-archives-xlviii-1-2024-129-2024","url":null,"abstract":"Abstract. In order to improve the data quality of land use remote sensing monitoring images, this article introduces the process of generating satellite image data, elaborates on the content of satellite image data verification in land use remote sensing monitoring, and proposes quality issues and improvement measures for satellite image data. Taking the discovered satellite image data quality issues as an example, compared with quality inspection standards, it was found that the main problems in the results were projection parameter errors, image color distortion, image blurring, and position accuracy exceeding limits. It is recommended to check the above issues during the image production stage, analyze the reasons for exceeding the position accuracy limit, image distortion, and embossing, and provide relevant suggestions. Provided strong technical support for land use surveys.\u0000","PeriodicalId":505918,"journal":{"name":"The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences","volume":" 3","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140992345","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 : 2024-05-10DOI: 10.5194/isprs-archives-xlviii-1-2024-37-2024
Yuezhen Cai, Linyuan Xia, Ting On Chan
Abstract. Calibrating optical sensors with common targets facilitates the efficient and convenient acquisition of the sensor's internal parameters. In this paper, we present a new method of camera calibration utilizing a low-cost foamy cube, in a form of dice, which is based on the fact that arrangement of pip and cubical die surfaces is mutually orthogonal. Initially, each face and pips are identified through the color information on the die’s surfaces. Subsequently, the centers of pips are corrected using a circular projection model, and radial distortion coefficients are estimated based on centers’ one-to-one correspondences. After that, the tangent information between pairs of pips on orthogonal dice faces are utilized to compute vanishing points, leading to estimation of intrinsic parameters. Experimental results demonstrate that our method has similar effects compared to well-known checkerboard calibration method, reaching an average relative error of 2.43%, simplifying the calibration process in practical applications and showcasing good practicality and robustness.
{"title":"A Straightforward Camera Calibration Method Based on a Single Low-cost Cubical Target","authors":"Yuezhen Cai, Linyuan Xia, Ting On Chan","doi":"10.5194/isprs-archives-xlviii-1-2024-37-2024","DOIUrl":"https://doi.org/10.5194/isprs-archives-xlviii-1-2024-37-2024","url":null,"abstract":"Abstract. Calibrating optical sensors with common targets facilitates the efficient and convenient acquisition of the sensor's internal parameters. In this paper, we present a new method of camera calibration utilizing a low-cost foamy cube, in a form of dice, which is based on the fact that arrangement of pip and cubical die surfaces is mutually orthogonal. Initially, each face and pips are identified through the color information on the die’s surfaces. Subsequently, the centers of pips are corrected using a circular projection model, and radial distortion coefficients are estimated based on centers’ one-to-one correspondences. After that, the tangent information between pairs of pips on orthogonal dice faces are utilized to compute vanishing points, leading to estimation of intrinsic parameters. Experimental results demonstrate that our method has similar effects compared to well-known checkerboard calibration method, reaching an average relative error of 2.43%, simplifying the calibration process in practical applications and showcasing good practicality and robustness.\u0000","PeriodicalId":505918,"journal":{"name":"The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences","volume":" 33","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140990942","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 : 2024-05-10DOI: 10.5194/isprs-archives-xlviii-1-2024-159-2024
A. Gaboutchian, V. Knyaz, Anatoly Maximov, S. Zheltov
Abstract. Among different study techniques, which are used in dental research, measurements of teeth play one of the most important roles. However classical measurement techniques usually show consistent results on morphologically complete objects, while teeth can be found in different conditions, depending, for instance, on the natural wear degree. In order to increase the sample, which is especially important when findings are not abundant, and to improve the analytical part, alternative techniques have been proposed, among which cervical measurements are considered to be informative, especially taking into consideration morphological and methodological importance of the cervical area. Algorithms of automated digital measurement techniques also use the cervical area for providing stability of results. Visualisation of the tooth cervical margin and reconstruction of its projections can be achieved by two conventional imaging techniques, which are tomography (preferably high-resolution) and intra-oral confocal optical scanning. They both were used for obtaining 3D reconstruction of upper premolar taken from palaeoanthropological materials. In line with applying the same automated coordinate system setting algorithm to both types of reconstructions, contours of their enamel cervical margins were defined and their projections to horizontal plane were obtained and measured. Despite the fact that 3D reconstructions from different imaging sources technically can serve for running automated odontometry, measurement results, especially in comparative studies, should be handled with attention.
{"title":"Influence of 3D models choice on cervical outline measurements of teeth","authors":"A. Gaboutchian, V. Knyaz, Anatoly Maximov, S. Zheltov","doi":"10.5194/isprs-archives-xlviii-1-2024-159-2024","DOIUrl":"https://doi.org/10.5194/isprs-archives-xlviii-1-2024-159-2024","url":null,"abstract":"Abstract. Among different study techniques, which are used in dental research, measurements of teeth play one of the most important roles. However classical measurement techniques usually show consistent results on morphologically complete objects, while teeth can be found in different conditions, depending, for instance, on the natural wear degree. In order to increase the sample, which is especially important when findings are not abundant, and to improve the analytical part, alternative techniques have been proposed, among which cervical measurements are considered to be informative, especially taking into consideration morphological and methodological importance of the cervical area. Algorithms of automated digital measurement techniques also use the cervical area for providing stability of results. Visualisation of the tooth cervical margin and reconstruction of its projections can be achieved by two conventional imaging techniques, which are tomography (preferably high-resolution) and intra-oral confocal optical scanning. They both were used for obtaining 3D reconstruction of upper premolar taken from palaeoanthropological materials. In line with applying the same automated coordinate system setting algorithm to both types of reconstructions, contours of their enamel cervical margins were defined and their projections to horizontal plane were obtained and measured. Despite the fact that 3D reconstructions from different imaging sources technically can serve for running automated odontometry, measurement results, especially in comparative studies, should be handled with attention.\u0000","PeriodicalId":505918,"journal":{"name":"The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences","volume":" 10","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140990207","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 : 2024-05-10DOI: 10.5194/isprs-archives-xlviii-1-2024-335-2024
Minglei Li, Li Xu, Mingfan Li, Guoyuan Qu, Dazhou Wei, Wei Li
Abstract. This paper proposes a method for identifying 3D point cloud of transmission line acquired by light detection and ranging (LiDAR) real-time mobile scanning. Since the single frame of point cloud obtained by LiDAR is sparse, the method employs a sliding spatial window strategy with Kalman filtering for dynamic point cloud registration. Then, a 3D point cloud deep learning neural network that utilizes uniform sampling and local feature aggregation (LFA) is designed specifically for transmission line objects. The network handles the problem of long-span objects and a large amount of point cloud. Finally, the instantiated transmission line objects are extracted from the top-down projection of the semantically segmented 3D point cloud by fast Euclidean clustering algorithm. Experiments demonstrate that the method achieves a classification accuracy of 94.7% and a mean intersection over union of 81.6% on 3D point cloud datasets of transmission line obtained from LiDAR mobile scanning, validating its ability to achieve real-time identification and distance measurement of transmission line objects.
{"title":"Mobile LiDAR-based Real-time Identification of Transmission Lines","authors":"Minglei Li, Li Xu, Mingfan Li, Guoyuan Qu, Dazhou Wei, Wei Li","doi":"10.5194/isprs-archives-xlviii-1-2024-335-2024","DOIUrl":"https://doi.org/10.5194/isprs-archives-xlviii-1-2024-335-2024","url":null,"abstract":"Abstract. This paper proposes a method for identifying 3D point cloud of transmission line acquired by light detection and ranging (LiDAR) real-time mobile scanning. Since the single frame of point cloud obtained by LiDAR is sparse, the method employs a sliding spatial window strategy with Kalman filtering for dynamic point cloud registration. Then, a 3D point cloud deep learning neural network that utilizes uniform sampling and local feature aggregation (LFA) is designed specifically for transmission line objects. The network handles the problem of long-span objects and a large amount of point cloud. Finally, the instantiated transmission line objects are extracted from the top-down projection of the semantically segmented 3D point cloud by fast Euclidean clustering algorithm. Experiments demonstrate that the method achieves a classification accuracy of 94.7% and a mean intersection over union of 81.6% on 3D point cloud datasets of transmission line obtained from LiDAR mobile scanning, validating its ability to achieve real-time identification and distance measurement of transmission line objects.\u0000","PeriodicalId":505918,"journal":{"name":"The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences","volume":" 31","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140993982","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 : 2024-05-10DOI: 10.5194/isprs-archives-xlviii-1-2024-555-2024
S. Pandit, S. Shimada, Timothy Dube
Abstract. This study aims to integrate driver variables with a land use change model (LCM) to explore their impact on the natural environment within the context of land-use changes in the Republic of Djibouti, considering possible Business-as-usual scenarios. Secondary data from 1990 and 2012 on land use land cover (LULC) were analyzed, with a 2022 map generated by adopting the same method of secondary data used (random forest classification in Google Earth Engine (GEE)) for validation. Eight key driver variables were utilized to model plausible future land cover (2035) for Djibouti. Statistical outputs and change maps from the LCM were compared to gauge historical change estimates and simulated scenarios. Analysis from 1990 to 2022 highlights significant land use and cover changes spurred by urbanization, environmental factors, and economic development. Barren land and bushland dominated, while built-up areas and water bodies expanded notably. Urbanization, agriculture, and climate change contributed to vegetation degradation, with declines in mangroves and increases in built-up areas. Water bodies also expanded during this period. Projections from the 2035 LULC map anticipate further urban expansion, underscoring the need for sustainable land management practices. In conclusion, comprehensive land-use planning, interdisciplinary approaches, and stakeholder engagement are deemed critical for addressing Djibouti's socio-economic and environmental challenges and steering towards a sustainable future. These simulated results offer valuable insights for regional governments to frame strategic policies and assess management actions for resource utilization amidst urbanization and population growth trends.
{"title":"Selected Driver Variables for the Simulation of Land-Use and Land-Cover Change for the Republic of Djibouti: A Study from Semi-Arid Region","authors":"S. Pandit, S. Shimada, Timothy Dube","doi":"10.5194/isprs-archives-xlviii-1-2024-555-2024","DOIUrl":"https://doi.org/10.5194/isprs-archives-xlviii-1-2024-555-2024","url":null,"abstract":"Abstract. This study aims to integrate driver variables with a land use change model (LCM) to explore their impact on the natural environment within the context of land-use changes in the Republic of Djibouti, considering possible Business-as-usual scenarios. Secondary data from 1990 and 2012 on land use land cover (LULC) were analyzed, with a 2022 map generated by adopting the same method of secondary data used (random forest classification in Google Earth Engine (GEE)) for validation. Eight key driver variables were utilized to model plausible future land cover (2035) for Djibouti. Statistical outputs and change maps from the LCM were compared to gauge historical change estimates and simulated scenarios. Analysis from 1990 to 2022 highlights significant land use and cover changes spurred by urbanization, environmental factors, and economic development. Barren land and bushland dominated, while built-up areas and water bodies expanded notably. Urbanization, agriculture, and climate change contributed to vegetation degradation, with declines in mangroves and increases in built-up areas. Water bodies also expanded during this period. Projections from the 2035 LULC map anticipate further urban expansion, underscoring the need for sustainable land management practices. In conclusion, comprehensive land-use planning, interdisciplinary approaches, and stakeholder engagement are deemed critical for addressing Djibouti's socio-economic and environmental challenges and steering towards a sustainable future. These simulated results offer valuable insights for regional governments to frame strategic policies and assess management actions for resource utilization amidst urbanization and population growth trends.\u0000","PeriodicalId":505918,"journal":{"name":"The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences","volume":" 2","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140991743","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 : 2024-05-10DOI: 10.5194/isprs-archives-xlviii-1-2024-291-2024
V. Knyaz, V. Kniaz, S. Zheltov, Kirill S. Petrov
Abstract. One of the urgent and constantly in demand problems is updating maps. Maps, representing geo-information in vector form, have undoubted advantages in compactness and ”readability” compared to aerial photographs. The issue of maps actuality is critically important for rational urban planning, precision farming, the relevance of the cadastre and other geospatial applications. Various sources of data are used for maps updating, with aerial imagery being the main and rich source of information. Automatic processing of aerial photographs makes it possible to efficiently extract vector information, providing operational monitoring and accounting for changes that have appeared. The presented study addresses the problem of multi sensor information fusion in order to obtain accurate vector information. We use aerial images as a main data source and additionally the data of laser scanning and ground survey to increase performance of automatic image semantic segmentation and vectorization. The proposed framework is demonstrated on the task of forest monitoring.
{"title":"Multi-sensor Data Analysis for Aerial Image Semantic Segmentation and Vectorization","authors":"V. Knyaz, V. Kniaz, S. Zheltov, Kirill S. Petrov","doi":"10.5194/isprs-archives-xlviii-1-2024-291-2024","DOIUrl":"https://doi.org/10.5194/isprs-archives-xlviii-1-2024-291-2024","url":null,"abstract":"Abstract. One of the urgent and constantly in demand problems is updating maps. Maps, representing geo-information in vector form, have undoubted advantages in compactness and ”readability” compared to aerial photographs. The issue of maps actuality is critically important for rational urban planning, precision farming, the relevance of the cadastre and other geospatial applications. Various sources of data are used for maps updating, with aerial imagery being the main and rich source of information. Automatic processing of aerial photographs makes it possible to efficiently extract vector information, providing operational monitoring and accounting for changes that have appeared. The presented study addresses the problem of multi sensor information fusion in order to obtain accurate vector information. We use aerial images as a main data source and additionally the data of laser scanning and ground survey to increase performance of automatic image semantic segmentation and vectorization. The proposed framework is demonstrated on the task of forest monitoring.\u0000","PeriodicalId":505918,"journal":{"name":"The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences","volume":" 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140992453","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 : 2024-05-10DOI: 10.5194/isprs-archives-xlviii-1-2024-679-2024
Xiangsheng Wang, Xikun Hu, Ping Zhong
Abstract. Object detection is a widely studied task in computer vision. Current methods often focus on images captured from appropriate viewpoints. However, there is a large disparity between objects observed from different viewpoints in the real world. Dynamic Object Detection (DOD) method automatically adjusts the camera viewpoint in a visual scene to sequentially find optimal viewpoints. Currently, the DOD tasks are usually modeled as a sequential decision-making problem and solved using reinforcement learning methods. Existing approaches face challenges with sparse rewards and training instability. To tackle these issues, we proposed a single-step reward function and a lightweight network, respectively. The single-step reward function, which provides timely feedback, gives an efficient training process for DOD tasks. The lightweight network with few parameters can ensure the stability of the training process. To evaluate the effectiveness of our method, we developed a simulation dataset based on UE4, which consists of 1800 training images and 450 testing images. The dataset includes five object categories: vans, cars, trailers, box trucks and SUVs. Experiments demonstrate that our method outperforms SOTA object detectors on our simulation dataset. Specifically, the average precisions(APs) are improved from 89.1% to 96.0% when using the YOLOv8 object detector.
摘要物体检测是计算机视觉中一项被广泛研究的任务。目前的方法通常侧重于从适当的视角捕捉图像。然而，在现实世界中，从不同视角观察到的物体之间存在很大差异。动态物体检测（DOD）方法可自动调整视觉场景中的摄像机视点，从而依次找到最佳视点。目前，DOD 任务通常被建模为一个顺序决策问题，并使用强化学习方法来解决。现有方法面临着奖励稀疏和训练不稳定的挑战。针对这些问题，我们分别提出了单步奖励函数和轻量级网络。单步奖励函数能提供及时反馈，为 DOD 任务提供了高效的训练过程。参数较少的轻量级网络可以确保训练过程的稳定性。为了评估我们方法的有效性，我们开发了一个基于 UE4 的模拟数据集，其中包括 1800 张训练图像和 450 张测试图像。该数据集包括五个对象类别：货车、轿车、拖车、箱式卡车和越野车。实验证明，在模拟数据集上，我们的方法优于 SOTA 物体检测器。具体来说，使用 YOLOv8 物体检测器时，平均精确度（APs）从 89.1% 提高到 96.0%。
{"title":"Visual Reinforcement Learning for Dynamic Object Detection","authors":"Xiangsheng Wang, Xikun Hu, Ping Zhong","doi":"10.5194/isprs-archives-xlviii-1-2024-679-2024","DOIUrl":"https://doi.org/10.5194/isprs-archives-xlviii-1-2024-679-2024","url":null,"abstract":"Abstract. Object detection is a widely studied task in computer vision. Current methods often focus on images captured from appropriate viewpoints. However, there is a large disparity between objects observed from different viewpoints in the real world. Dynamic Object Detection (DOD) method automatically adjusts the camera viewpoint in a visual scene to sequentially find optimal viewpoints. Currently, the DOD tasks are usually modeled as a sequential decision-making problem and solved using reinforcement learning methods. Existing approaches face challenges with sparse rewards and training instability. To tackle these issues, we proposed a single-step reward function and a lightweight network, respectively. The single-step reward function, which provides timely feedback, gives an efficient training process for DOD tasks. The lightweight network with few parameters can ensure the stability of the training process. To evaluate the effectiveness of our method, we developed a simulation dataset based on UE4, which consists of 1800 training images and 450 testing images. The dataset includes five object categories: vans, cars, trailers, box trucks and SUVs. Experiments demonstrate that our method outperforms SOTA object detectors on our simulation dataset. Specifically, the average precisions(APs) are improved from 89.1% to 96.0% when using the YOLOv8 object detector.\u0000","PeriodicalId":505918,"journal":{"name":"The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences","volume":" 7","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140991058","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：