Pub Date : 2023-06-06DOI: 10.5194/agile-giss-4-50-2023
Yiyu Wang, Jiaqi Ge, A. Comber
Abstract. This research proposed an improved pedestrian evacuation ABM incorporating Bayesian Nash equilibrium (BNE) to provide more realistic simulations of evacuating behaviours in complex environments. BNE theory was introduced to improve the rationality of model simulations by quantifying individual decision-making process. Latest research put forward that BNE pedestrians (agents) were capable of evacuating faster and displayed more intelligent and representative evacuating behaviours. To further evaluate the role of BNE played in agents’ navigations in complex scenarios, this paper extends the above work by introducing impassable barriers with changeable sizes to realise the simulations in a more complex evacuation space with several narrow corridors. In order to match the demands of efficiently avoiding congestions and impassable areas, the decision-making rule of BNE agents when one patch was occupied by over 10 agents was improved from 100% best strategy to a multi-strategy combination: with 50% optimal strategy, 40% suboptimal strategy and 10% choosing one of the remaining options. It was found that compared with the agents following the other two traditional models, BNE agents could change their original exiting route after considering possible movements of the neighbouring agents and may evacuate through the corridors relatively further from the exit. A detailed introduction of the improved ABM is provided in this paper. Potential research directions are also identified.
{"title":"A pedestrian ABM in complex evacuation environments based on Bayesian Nash Equilibrium","authors":"Yiyu Wang, Jiaqi Ge, A. Comber","doi":"10.5194/agile-giss-4-50-2023","DOIUrl":"https://doi.org/10.5194/agile-giss-4-50-2023","url":null,"abstract":"Abstract. This research proposed an improved pedestrian evacuation ABM incorporating Bayesian Nash equilibrium (BNE) to provide more realistic simulations of evacuating behaviours in complex environments. BNE theory was introduced to improve the rationality of model simulations by quantifying individual decision-making process. Latest research put forward that BNE pedestrians (agents) were capable of evacuating faster and displayed more intelligent and representative evacuating behaviours. To further evaluate the role of BNE played in agents’ navigations in complex scenarios, this paper extends the above work by introducing impassable barriers with changeable sizes to realise the simulations in a more complex evacuation space with several narrow corridors. In order to match the demands of efficiently avoiding congestions and impassable areas, the decision-making rule of BNE agents when one patch was occupied by over 10 agents was improved from 100% best strategy to a multi-strategy combination: with 50% optimal strategy, 40% suboptimal strategy and 10% choosing one of the remaining options. It was found that compared with the agents following the other two traditional models, BNE agents could change their original exiting route after considering possible movements of the neighbouring agents and may evacuate through the corridors relatively further from the exit. A detailed introduction of the improved ABM is provided in this paper. Potential research directions are also identified.\u0000","PeriodicalId":116168,"journal":{"name":"AGILE: GIScience Series","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115446815","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 : 2023-06-06DOI: 10.5194/agile-giss-4-8-2023
Vasileios Milias, Shahin Sharifi Noorian, A. Bozzon, A. Psyllidis
Abstract. City streets that feel safe and attractive motivate active travel behaviour and promote people’s well-being. However, determining what makes a street safe and attractive is a challenging task because subjective qualities of the streetscape are difficult to quantify. Existing evidence typically focuses on how different street features influence perceived safety or attractiveness, but little is known about what influences both. To fill this knowledge gap, we developed a crowdsourcing tool and conducted a study with 403 participants, who were asked to virtually navigate city streets in Frankfurt, Germany, through a sequence of street-level images, rate locations based on perceived safety and attractiveness, and explain their ratings. Our results contribute new insights regarding the key similarities and differences between the factors influencing perceived safety and attractiveness. We show that the presence of human activity is strongly related to perceived safety, whereas attractiveness is influenced primarily by aesthetic qualities, as well as the number and type of amenities along a street. Moreover, we demonstrate that the presence of construction sites and underpasses has a disproportionately negative impact on perceived safety and attractiveness, outweighing the influence of any other features. We use the results to make evidence-informed recommendations for designing safer and more attractive streets that encourage active travel modes and promote well-being.
{"title":"Is it safe to be attractive? Disentangling the influence of streetscape features on the perceived safety and attractiveness of city streets","authors":"Vasileios Milias, Shahin Sharifi Noorian, A. Bozzon, A. Psyllidis","doi":"10.5194/agile-giss-4-8-2023","DOIUrl":"https://doi.org/10.5194/agile-giss-4-8-2023","url":null,"abstract":"Abstract. City streets that feel safe and attractive motivate active travel behaviour and promote people’s well-being. However, determining what makes a street safe and attractive is a challenging task because subjective qualities of the streetscape are difficult to quantify. Existing evidence typically focuses on how different street features influence perceived safety or attractiveness, but little is known about what influences both. To fill this knowledge gap, we developed a crowdsourcing tool and conducted a study with 403 participants, who were asked to virtually navigate city streets in Frankfurt, Germany, through a sequence of street-level images, rate locations based on perceived safety and attractiveness, and explain their ratings. Our results contribute new insights regarding the key similarities and differences between the factors influencing perceived safety and attractiveness. We show that the presence of human activity is strongly related to perceived safety, whereas attractiveness is influenced primarily by aesthetic qualities, as well as the number and type of amenities along a street. Moreover, we demonstrate that the presence of construction sites and underpasses has a disproportionately negative impact on perceived safety and attractiveness, outweighing the influence of any other features. We use the results to make evidence-informed recommendations for designing safer and more attractive streets that encourage active travel modes and promote well-being. \u0000","PeriodicalId":116168,"journal":{"name":"AGILE: GIScience Series","volume":"45 21","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113939015","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 : 2022-06-24DOI: 10.5194/agile-giss-3-72-2022
Francisco Garrido-Valenzuela, Sander van Cranenburgh, O. Cats
Abstract. Characteristics of urban space (co-)determine human behaviour, including their social interaction patterns. However, despite numerous studies that have examined how the urban space impacts social interactions, their relationships are still poorly understood. Recent developments in computer vision and machine learning fields offer promising new ways to analyse and collect data on social interactions. This study proposes a new computer vision-based approach to study how the urban space impacts social interactions. The proposed method uses pre-trained object detection models to detect social interactions (including their geo-locations) from street-view imagery. After that, it regresses urban space characteristics – which are also detected using object detection models – on social interactions. For this study, 294,852 street-level images from three Dutch cities are analysed. Results from linear regression analysis show that for these three Dutch cities people tend to meet in places with a strong presence of recreational attractions and bicycles. Also, the results of data collection and image processing can be used to identify the areas most likely to produce social interactions in urban space to conduct urban studies.
{"title":"Enriching geospatial data with computer vision to identify urban environment determinants of social interactions","authors":"Francisco Garrido-Valenzuela, Sander van Cranenburgh, O. Cats","doi":"10.5194/agile-giss-3-72-2022","DOIUrl":"https://doi.org/10.5194/agile-giss-3-72-2022","url":null,"abstract":"Abstract. Characteristics of urban space (co-)determine human behaviour, including their social interaction patterns. However, despite numerous studies that have examined how the urban space impacts social interactions, their relationships are still poorly understood. Recent developments in computer vision and machine learning fields offer promising new ways to analyse and collect data on social interactions. This study proposes a new computer vision-based approach to study how the urban space impacts social interactions. The proposed method uses pre-trained object detection models to detect social interactions (including their geo-locations) from street-view imagery. After that, it regresses urban space characteristics – which are also detected using object detection models – on social interactions. For this study, 294,852 street-level images from three Dutch cities are analysed. Results from linear regression analysis show that for these three Dutch cities people tend to meet in places with a strong presence of recreational attractions and bicycles. Also, the results of data collection and image processing can be used to identify the areas most likely to produce social interactions in urban space to conduct urban studies.\u0000","PeriodicalId":116168,"journal":{"name":"AGILE: GIScience Series","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114571495","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 : 2022-06-11DOI: 10.5194/agile-giss-3-43-2022
Christian Koski, P. Kettunen, Justus Poutanen, J. Oksanen
Abstract. Deep learning methods for semantic segmentation have shown great potential in automating mapping of geospatial features, including small watercourses such as streams and ditches. There are a variety of small watercourse types. In many use cases users are only interested in specific types of watercourses. However, the impact on results from neural networks trained with only some types of small watercourses, compared to all types of watercourses is not well known. We trained four deep learning models to semantically segment watercourses from an elevation model. One model was trained with all small watercourses in the labels as a single class, while three models were trained each with a single type of watercourse in the label data. The results show that training the network with a single type of watercourse results in worse recall for all three watercourse types, compared to when training all of them together. This indicates that if the goal is to get as complete set of features as possible, it is better to include all watercourse types in the training data. Future studies could use multi-class output from neural network to determine how well networks could automatically classify features when training with all small watercourses in an area.
{"title":"Mapping small watercourses with deep learning – impact of training watercourse types separately","authors":"Christian Koski, P. Kettunen, Justus Poutanen, J. Oksanen","doi":"10.5194/agile-giss-3-43-2022","DOIUrl":"https://doi.org/10.5194/agile-giss-3-43-2022","url":null,"abstract":"Abstract. Deep learning methods for semantic segmentation have shown great potential in automating mapping of geospatial features, including small watercourses such as streams and ditches. There are a variety of small watercourse types. In many use cases users are only interested in specific types of watercourses. However, the impact on results from neural networks trained with only some types of small watercourses, compared to all types of watercourses is not well known. We trained four deep learning models to semantically segment watercourses from an elevation model. One model was trained with all small watercourses in the labels as a single class, while three models were trained each with a single type of watercourse in the label data. The results show that training the network with a single type of watercourse results in worse recall for all three watercourse types, compared to when training all of them together. This indicates that if the goal is to get as complete set of features as possible, it is better to include all watercourse types in the training data. Future studies could use multi-class output from neural network to determine how well networks could automatically classify features when training with all small watercourses in an area.\u0000","PeriodicalId":116168,"journal":{"name":"AGILE: GIScience Series","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130421507","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 : 2022-06-11DOI: 10.5194/agile-giss-3-61-2022
Lucas Spierenburg, Sander van Cranenburgh, O. Cats
Abstract. Regionalization is the process of aggregating contiguous spatial units to form areas that are homogeneous with respect to one or a set of variables. It is useful when studying spatial phenomena or when designing region-based policies, as it allows to unravel the latent spatial structure of a dataset. However, this task is challenging when small-scale fluctuations in the data interfere with the phenomenon of interest. In such circumstances, regionalization techniques are prone to overfitting small-scale fluctuations, and producing erratic regions. This paper presents a regionalization method robust to small-scale variations that is particularly relevant when handling demographic data. Fluctuations are filtered out using a weighted spatial average before applying agglomerative clustering. The method is tested against a conventional agglomerative clustering approach on a fine-resolution demographic dataset, for a set of indicators quantifying: the ability to identify large-scale spatial patterns, the homogeneity of the regions produced, and the spatial regularity of these regions. These indicators have been computed for the two methods for a number of clusters ranging from 2 to 101, and results show that the proposed approach performs better than conventional agglomerative clustering more than 90% of the time at identifying large-scale patterns, and produces more regular regions 96% of the time.
{"title":"A regionalization method filtering out small-scale spatial fluctuations","authors":"Lucas Spierenburg, Sander van Cranenburgh, O. Cats","doi":"10.5194/agile-giss-3-61-2022","DOIUrl":"https://doi.org/10.5194/agile-giss-3-61-2022","url":null,"abstract":"Abstract. Regionalization is the process of aggregating contiguous spatial units to form areas that are homogeneous with respect to one or a set of variables. It is useful when studying spatial phenomena or when designing region-based policies, as it allows to unravel the latent spatial structure of a dataset. However, this task is challenging when small-scale fluctuations in the data interfere with the phenomenon of interest. In such circumstances, regionalization techniques are prone to overfitting small-scale fluctuations, and producing erratic regions. This paper presents a regionalization method robust to small-scale variations that is particularly relevant when handling demographic data. Fluctuations are filtered out using a weighted spatial average before applying agglomerative clustering. The method is tested against a conventional agglomerative clustering approach on a fine-resolution demographic dataset, for a set of indicators quantifying: the ability to identify large-scale spatial patterns, the homogeneity of the regions produced, and the spatial regularity of these regions. These indicators have been computed for the two methods for a number of clusters ranging from 2 to 101, and results show that the proposed approach performs better than conventional agglomerative clustering more than 90% of the time at identifying large-scale patterns, and produces more regular regions 96% of the time.\u0000","PeriodicalId":116168,"journal":{"name":"AGILE: GIScience Series","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124768741","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 : 2022-06-10DOI: 10.5194/agile-giss-3-5-2022
Marina Georgati, J. Monteiro, Bruno Martins, C. Kessler
Abstract. Accurate and consistent estimations on the present and future population distribution, at fine spatial resolution, are fundamental to support a variety of activities. However, the sampling regime, sample size, and methods used to collect census data are heterogeneous across temporal periods and/or geographic regions. Moreover, the data is usually only made available in aggregated form, to ensure privacy. In an attempt to address these issues, several previous initiatives have addressed the use of spatial disaggregation methods to produce high-resolution gridded datasets describing the human population distribution, although these projects have usually not addressed specific population subgroups. This paper describes a spatial disaggregation method based on self-training regression models, innovating over previous studies in the simultaneous prediction of disaggregated counts for multiple inter-related variables, by leveraging multi-output models based on gradient tree boosting. We report on experiments for two case studies, using high-resolution data (i.e., counts for different subgroups available at a resolution of 100 meters) for the municipality of Amsterdam and the region of Greater Copenhagen. Results show that the proposed approach can capture spatial heterogeneity and the dependency on local factors, outperforming alternatives (e.g., seminal disaggregation algorithms, or approaches leveraging individual regression models for each variable) in terms of averaged error metrics, and also upon visual inspection of spatial variation in the resulting maps.
{"title":"Spatial Disaggregation of Population Subgroups Leveraging Self-Trained Multi-Output Gradient Boosting Regression Trees","authors":"Marina Georgati, J. Monteiro, Bruno Martins, C. Kessler","doi":"10.5194/agile-giss-3-5-2022","DOIUrl":"https://doi.org/10.5194/agile-giss-3-5-2022","url":null,"abstract":"Abstract. Accurate and consistent estimations on the present and future population distribution, at fine spatial resolution, are fundamental to support a variety of activities. However, the sampling regime, sample size, and methods used to collect census data are heterogeneous across temporal periods and/or geographic regions. Moreover, the data is usually only made available in aggregated form, to ensure privacy. In an attempt to address these issues, several previous initiatives have addressed the use of spatial disaggregation methods to produce high-resolution gridded datasets describing the human population distribution, although these projects have usually not addressed specific population subgroups. This paper describes a spatial disaggregation method based on self-training regression models, innovating over previous studies in the simultaneous prediction of disaggregated counts for multiple inter-related variables, by leveraging multi-output models based on gradient tree boosting. We report on experiments for two case studies, using high-resolution data (i.e., counts for different subgroups available at a resolution of 100 meters) for the municipality of Amsterdam and the region of Greater Copenhagen. Results show that the proposed approach can capture spatial heterogeneity and the dependency on local factors, outperforming alternatives (e.g., seminal disaggregation algorithms, or approaches leveraging individual regression models for each variable) in terms of averaged error metrics, and also upon visual inspection of spatial variation in the resulting maps.\u0000","PeriodicalId":116168,"journal":{"name":"AGILE: GIScience Series","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129365642","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 : 2022-06-10DOI: 10.5194/agile-giss-3-31-2022
A. Comber, P. Harris, C. Brunsdon
Abstract. This paper describes initial work exploring GAM Gaussian Process (GP) splines parameterised by observation location, as a geographical varying coefficient model. Similar to GWR, this approach accommodates process spatial heterogeneity and generates spatially distributed, local coefficient estimates. These can be mapped to indicate the nature of the heterogeneity. The paper investigates the effect of the smoothing parameters used in the splines and how they alter the nature of the modelled heterogeneity. It optimises these in the GAM GP and the tuned model has subtle but important differences with the initial model. This has impacts on the nature of the process understanding (inference) that can be extracted from the model. This in turn suggest the need examine the underlying semantics of the resultant models in relation to the scale of process suggested by the smoothing parameters. A number of areas of further work are identified.
{"title":"Spatially Varying Coefficient Regression with GAM Gaussian Process splines: GAM(e)-on","authors":"A. Comber, P. Harris, C. Brunsdon","doi":"10.5194/agile-giss-3-31-2022","DOIUrl":"https://doi.org/10.5194/agile-giss-3-31-2022","url":null,"abstract":"Abstract. This paper describes initial work exploring GAM Gaussian Process (GP) splines parameterised by observation location, as a geographical varying coefficient model. Similar to GWR, this approach accommodates process spatial heterogeneity and generates spatially distributed, local coefficient estimates. These can be mapped to indicate the nature of the heterogeneity. The paper investigates the effect of the smoothing parameters used in the splines and how they alter the nature of the modelled heterogeneity. It optimises these in the GAM GP and the tuned model has subtle but important differences with the initial model. This has impacts on the nature of the process understanding (inference) that can be extracted from the model. This in turn suggest the need examine the underlying semantics of the resultant models in relation to the scale of process suggested by the smoothing parameters. A number of areas of further work are identified.\u0000","PeriodicalId":116168,"journal":{"name":"AGILE: GIScience Series","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129028492","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 : 2022-06-10DOI: 10.5194/agile-giss-3-34-2022
Heiko Figgemeier, Arne Rümmler, Christin Henzen
Abstract. In today’s research data management, experts discuss datasets to be FAIR, as they should become findable, accessible, interoperable and reusable (Lacagnia et al. 2021). In recent years, quality information and provenance information as well as dataset’s general metadata have become important aspects to evaluate a dataset's fitness for use. In order to capture and process this meta-information in a systematic way, users need frameworks and meaningful user interfaces that allow them to interact with the information and to visualize them. Therefore, we provide a user-friendly and interactive geodashboard implementation as first prototype that supports the evaluation of spatial datasets with linked widgets by applying semantic concepts and using open source libraries.
摘要在今天的研究数据管理中，专家们讨论的数据集应该是公平的，因为它们应该变得可查找、可访问、可互操作和可重用(Lacagnia et al. 2021)。近年来，质量信息和来源信息以及数据集的一般元数据已成为评估数据集适用性的重要方面。为了以系统的方式捕获和处理这些元信息，用户需要框架和有意义的用户界面，允许他们与信息交互并将其可视化。因此，我们提供了一个用户友好的交互式地理指示板实现，作为第一个原型，通过应用语义概念和使用开源库，支持使用链接小部件对空间数据集进行评估。
{"title":"A Geospatial Dashboard Prototype for Evaluating Spatial Datasets by using Semantic Data Concepts and Open Source Libraries","authors":"Heiko Figgemeier, Arne Rümmler, Christin Henzen","doi":"10.5194/agile-giss-3-34-2022","DOIUrl":"https://doi.org/10.5194/agile-giss-3-34-2022","url":null,"abstract":"Abstract. In today’s research data management, experts discuss datasets to be FAIR, as they should become findable, accessible, interoperable and reusable (Lacagnia et al. 2021). In recent years, quality information and provenance information as well as dataset’s general metadata have become important aspects to evaluate a dataset's fitness for use. In order to capture and process this meta-information in a systematic way, users need frameworks and meaningful user interfaces that allow them to interact with the information and to visualize them. Therefore, we provide a user-friendly and interactive geodashboard implementation as first prototype that supports the evaluation of spatial datasets with linked widgets by applying semantic concepts and using open source libraries.\u0000","PeriodicalId":116168,"journal":{"name":"AGILE: GIScience Series","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126762419","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 : 2022-06-10DOI: 10.5194/agile-giss-3-2-2022
Guoray Cai, Yimu Pan
Abstract. Point clouds data provides new potentials for automated construction of more geometrically accurate and semantically rich 3D models for indoor environments. Recent advances in deep learning methods on point cloud semantic segmentation demonstrated impressive accuracy in labeling points of 3D surfaces with object classes. However, it remains challenging to reconstruct the shape of semantic objects from semantically-labeled 3D points, due to imperfection of such data and the under-determination of object construction algorithms. We have little empirical knowledge about how data imperfections affect the reconstruction of 3D indoor room objects. This paper contributes to understanding the nature of such imperfection of 3D point cloud data and semantic segmentation algorithms by analyzing the reconstructability of indoor room objects from semantically-labeled point cloud. 181 rooms from Stanford Large-Scale 3D Indoor Spaces Dataset (S3DIS) were used in our experiment. After generating semantic labels on point-clouds using PointNet++ segmentic segmentation algorithm, we use human coders to judge the reconstructability of indoor objects, following a qualitative coding scheme. Human exploration of object shape imperfection was assisted by a visual analytic tool in making their judgement. We found that high point-level accuracy achieved through semantic segmentation of point cloud data does not guarantee high object-level accuracy. The extent of this problem varies widely among different spatial settings and configurations. We discuss the significance of these findings on the choice of 3D reconstruction methods.
{"title":"Understanding the Imperfection of 3D point Cloud and Semantic Segmentation algorithms for 3D Models of Indoor Environment","authors":"Guoray Cai, Yimu Pan","doi":"10.5194/agile-giss-3-2-2022","DOIUrl":"https://doi.org/10.5194/agile-giss-3-2-2022","url":null,"abstract":"Abstract. Point clouds data provides new potentials for automated construction of more geometrically accurate and semantically rich 3D models for indoor environments. Recent advances in deep learning methods on point cloud semantic segmentation demonstrated impressive accuracy in labeling points of 3D surfaces with object classes. However, it remains challenging to reconstruct the shape of semantic objects from semantically-labeled 3D points, due to imperfection of such data and the under-determination of object construction algorithms. We have little empirical knowledge about how data imperfections affect the reconstruction of 3D indoor room objects. This paper contributes to understanding the nature of such imperfection of 3D point cloud data and semantic segmentation algorithms by analyzing the reconstructability of indoor room objects from semantically-labeled point cloud. 181 rooms from Stanford Large-Scale 3D Indoor Spaces Dataset (S3DIS) were used in our experiment. After generating semantic labels on point-clouds using PointNet++ segmentic segmentation algorithm, we use human coders to judge the reconstructability of indoor objects, following a qualitative coding scheme. Human exploration of object shape imperfection was assisted by a visual analytic tool in making their judgement. We found that high point-level accuracy achieved through semantic segmentation of point cloud data does not guarantee high object-level accuracy. The extent of this problem varies widely among different spatial settings and configurations. We discuss the significance of these findings on the choice of 3D reconstruction methods. \u0000","PeriodicalId":116168,"journal":{"name":"AGILE: GIScience Series","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121274694","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 : 2022-06-10DOI: 10.5194/agile-giss-3-30-2022
A. Comber, Paul Harris, Kristina Bratkova, H. H. Phe, M. Kieu, Quang Thanh Bui, Thi-Thuy-Nghiem Nguyen, Eric Wanjau, N. Malleson
Abstract. The Modifiable Areal Unit Problem or MAUP is frequently alluded to but rarely addressed directly. The MAUP posits that statistical distributions, relationships and trends can exhibit very different properties when the same data are aggregated or combined over different reporting units or scales. This paper explores a number of approaches for determining appropriate scales of spatial aggregation. It examines a travel survey, undertaken in Ha Noi, Vietnam, that captures attitudes towards a potential ban of motorised transport in the city centre. The data are rich, capturing travel destinations, purposes, modes and frequencies, as well as respondent demographics (age, occupation, housing etc) including home locations. The dataset is highly dimensional, with a large n (26339 records) and a large m (142 fields). When the raw individual level data are used to analyse the factors associated with travel ban attitudes, the resultant models are weak and inconclusive - the data are too noisy. Aggregating the data can overcome this, but this raises the question of appropriate aggregation scales. This paper demonstrates how aggregation scales can be evaluated using a range of different metrics related to spatial and non-spatial variances. In so doing it demonstrates how the MAUP can be directly addressed in analyses of spatial data.
{"title":"Handling the MAUP: methods for identifying appropriate scales of aggregation based on measures on spatial and non-spatial variance","authors":"A. Comber, Paul Harris, Kristina Bratkova, H. H. Phe, M. Kieu, Quang Thanh Bui, Thi-Thuy-Nghiem Nguyen, Eric Wanjau, N. Malleson","doi":"10.5194/agile-giss-3-30-2022","DOIUrl":"https://doi.org/10.5194/agile-giss-3-30-2022","url":null,"abstract":"Abstract. The Modifiable Areal Unit Problem or MAUP is frequently alluded to but rarely addressed directly. The MAUP posits that statistical distributions, relationships and trends can exhibit very different properties when the same data are aggregated or combined over different reporting units or scales. This paper explores a number of approaches for determining appropriate scales of spatial aggregation. It examines a travel survey, undertaken in Ha Noi, Vietnam, that captures attitudes towards a potential ban of motorised transport in the city centre. The data are rich, capturing travel destinations, purposes, modes and frequencies, as well as respondent demographics (age, occupation, housing etc) including home locations. The dataset is highly dimensional, with a large n (26339 records) and a large m (142 fields). When the raw individual level data are used to analyse the factors associated with travel ban attitudes, the resultant models are weak and inconclusive - the data are too noisy. Aggregating the data can overcome this, but this raises the question of appropriate aggregation scales. This paper demonstrates how aggregation scales can be evaluated using a range of different metrics related to spatial and non-spatial variances. In so doing it demonstrates how the MAUP can be directly addressed in analyses of spatial data.\u0000","PeriodicalId":116168,"journal":{"name":"AGILE: GIScience Series","volume":"418 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116683480","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：