Yun Si Goh, Jing Wen Leong, Seanglidet Yean, B. Lee, K. M. Ngo, Pete Edwards
While the benefits of trees are well-known, there are few studies on the vegetation cover in Singapore as traditional data acquisition is inefficient. In this study, we put together an efficient land use classification pipeline for the highly urbanized country using Sentinel-2 (S2) images. We adopted an object-based (OB) approach which uses Simple Non-iterative Clustering (SNIC) for clustering and Grey Level Co-occurrence Matrix (GLCM) for textural indices. Random Forest (RF) classifier was used for classification. We produced maps with 85.8% accuracy for the years 2016 to 2021. We then analysed the vegetation cover changes using change detection methods, and identified areas with significant vegetation loss (24.4km2 or 3.14% of our study area) or gain (40.4km2 or 5.20% of our study area). We also determined the type of land use conversions in these areas. This study contributes to tree management, environmental impact assessments (EIA) and policy-making. It also lays the groundwork for future studies on city livability.
{"title":"A study on Singapore's vegetation cover and land use change using remote sensing","authors":"Yun Si Goh, Jing Wen Leong, Seanglidet Yean, B. Lee, K. M. Ngo, Pete Edwards","doi":"10.1145/3557922.3567480","DOIUrl":"https://doi.org/10.1145/3557922.3567480","url":null,"abstract":"While the benefits of trees are well-known, there are few studies on the vegetation cover in Singapore as traditional data acquisition is inefficient. In this study, we put together an efficient land use classification pipeline for the highly urbanized country using Sentinel-2 (S2) images. We adopted an object-based (OB) approach which uses Simple Non-iterative Clustering (SNIC) for clustering and Grey Level Co-occurrence Matrix (GLCM) for textural indices. Random Forest (RF) classifier was used for classification. We produced maps with 85.8% accuracy for the years 2016 to 2021. We then analysed the vegetation cover changes using change detection methods, and identified areas with significant vegetation loss (24.4km2 or 3.14% of our study area) or gain (40.4km2 or 5.20% of our study area). We also determined the type of land use conversions in these areas. This study contributes to tree management, environmental impact assessments (EIA) and policy-making. It also lays the groundwork for future studies on city livability.","PeriodicalId":393750,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL International Workshop on Spatial Big Data and AI for Industrial Applications","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116775649","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}
Building numbers shown on building outlines of a map are important information for guiding delivery associates to the correct building of a package's recipient. Intuitively, the more labeled buildings are present in our map, the less likely to misplace an order in addition to other benefits such as delivery efficiency as drivers get better visual cues about building positions. Although there are free and collaborative projects for creating geographic database of the world, such as the OpenStreetMap (OSM) [2] which also supplies building outlines along with their building numbers, many building outlines still remain unlabeled in many U.S. regions and other countries. Hence, we are interested in developing models that can automatically add building numbers with ≥ 99% precision to unlabeled buildings across geographies with low to medium building number coverage. In this paper, we describe a ML model which in offline results showed 2% to 12% increase in building number coverage in some US regions compared to that of the OSM. The proposed model can also be applied to improve the building number coverage of other countries after fine-tuning to those new regions.
{"title":"BinoML","authors":"Puyang Ma, Ravi Garg, Mohamed M. A. Moustafa","doi":"10.1145/3557922.3567481","DOIUrl":"https://doi.org/10.1145/3557922.3567481","url":null,"abstract":"Building numbers shown on building outlines of a map are important information for guiding delivery associates to the correct building of a package's recipient. Intuitively, the more labeled buildings are present in our map, the less likely to misplace an order in addition to other benefits such as delivery efficiency as drivers get better visual cues about building positions. Although there are free and collaborative projects for creating geographic database of the world, such as the OpenStreetMap (OSM) [2] which also supplies building outlines along with their building numbers, many building outlines still remain unlabeled in many U.S. regions and other countries. Hence, we are interested in developing models that can automatically add building numbers with ≥ 99% precision to unlabeled buildings across geographies with low to medium building number coverage. In this paper, we describe a ML model which in offline results showed 2% to 12% increase in building number coverage in some US regions compared to that of the OSM. The proposed model can also be applied to improve the building number coverage of other countries after fine-tuning to those new regions.","PeriodicalId":393750,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL International Workshop on Spatial Big Data and AI for Industrial Applications","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117344914","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}
Ehsan Jalilifar, Xiao Li, Michael E. Martin, Xiao-Xiao Huang
Effectively monitoring border crossing time is of great importance to various stakeholders. Border crossing information systems currently implemented along the United States-Mexico border require a large installed base of sensors, costly for installation and maintenance. This study provides a preliminary assessment of market-available connected vehicle (CV) data in border crossing time estimation. We evaluated one week of CV data collected at the Paso del Norte (PDN) port of entry (POE). We used a set of big data analytic tools to process big CV datasets and generated CV-based border crossing times (CV-Time). Then, we evaluated the correlation between the CV-Time and the existing Bluetooth-generated border crossing times (Bluetooth-Time) at the PDN POE. Last, we built a regression model to estimate the Bluetooth-Time (ground truth data) based on CV-based variables. The results demonstrate that the CV-Time is strongly correlated with the Bluetooth-Time, with a correlation rate of approximately 0.89. This study demonstrates that the market-available CV data is a potential data source for monitoring border crossing times.
有效监测过境时间对各利益相关方都非常重要。目前沿美墨边境实施的过境信息系统需要安装大量传感器,安装和维护费用高昂。本研究提供了一个初步的评估市场上可用的互联车辆(CV)数据在过境时间估计。我们评估了在Paso del Norte (PDN)入境口岸(POE)收集的一周CV数据。我们使用一套大数据分析工具来处理大CV数据集,并生成基于CV的边界穿越时间(CV- time)。然后,我们评估了CV-Time与PDN POE中现有蓝牙生成的边界穿越次数(Bluetooth-Time)之间的相关性。最后,我们建立了一个回归模型来估计基于cv的变量的蓝牙时间(地面真实数据)。结果表明,cvtime与Bluetooth-Time具有很强的相关性,相关率约为0.89。本研究表明,市场上可获得的CV数据是监测过境次数的潜在数据源。
{"title":"Toward a crowdsourcing solution to estimate border crossing times using market-available connected vehicle data","authors":"Ehsan Jalilifar, Xiao Li, Michael E. Martin, Xiao-Xiao Huang","doi":"10.1145/3557922.3567669","DOIUrl":"https://doi.org/10.1145/3557922.3567669","url":null,"abstract":"Effectively monitoring border crossing time is of great importance to various stakeholders. Border crossing information systems currently implemented along the United States-Mexico border require a large installed base of sensors, costly for installation and maintenance. This study provides a preliminary assessment of market-available connected vehicle (CV) data in border crossing time estimation. We evaluated one week of CV data collected at the Paso del Norte (PDN) port of entry (POE). We used a set of big data analytic tools to process big CV datasets and generated CV-based border crossing times (CV-Time). Then, we evaluated the correlation between the CV-Time and the existing Bluetooth-generated border crossing times (Bluetooth-Time) at the PDN POE. Last, we built a regression model to estimate the Bluetooth-Time (ground truth data) based on CV-based variables. The results demonstrate that the CV-Time is strongly correlated with the Bluetooth-Time, with a correlation rate of approximately 0.89. This study demonstrates that the market-available CV data is a potential data source for monitoring border crossing times.","PeriodicalId":393750,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL International Workshop on Spatial Big Data and AI for Industrial Applications","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116239618","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}
{"title":"Proceedings of the 1st ACM SIGSPATIAL International Workshop on Spatial Big Data and AI for Industrial Applications","authors":"","doi":"10.1145/3557922","DOIUrl":"https://doi.org/10.1145/3557922","url":null,"abstract":"","PeriodicalId":393750,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL International Workshop on Spatial Big Data and AI for Industrial Applications","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133212607","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}