{"title":"GEO-TAGGED IMAGE RETRIEVAL FROM MAPILLARY STREET IMAGES FOR A TARGET BUILDING","authors":"N. Celik, E. Sümer","doi":"10.5194/isprs-archives-xliv-4-w3-2020-151-2020","DOIUrl":null,"url":null,"abstract":"Abstract. This study aims to investigate the possibility to automate the image selection process for the target building from Mapillary images through a web application where the user only initiates one image of the target building as a query. Using the data provided with Mapillary API and Overpass API, all images having full or partial coverage of the target building were selected. Then the images were segmented by using a pre-trained U-Net model to discard any images having less than 20% building coverage. The experiments showed promising results yielding 0.971 and 0.887 of overall accuracy after segmentation steps for two different target buildings.","PeriodicalId":14757,"journal":{"name":"ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences","volume":"106 1","pages":"151-158"},"PeriodicalIF":0.0000,"publicationDate":"2020-11-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.5194/isprs-archives-xliv-4-w3-2020-151-2020","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
Abstract
Abstract. This study aims to investigate the possibility to automate the image selection process for the target building from Mapillary images through a web application where the user only initiates one image of the target building as a query. Using the data provided with Mapillary API and Overpass API, all images having full or partial coverage of the target building were selected. Then the images were segmented by using a pre-trained U-Net model to discard any images having less than 20% building coverage. The experiments showed promising results yielding 0.971 and 0.887 of overall accuracy after segmentation steps for two different target buildings.
京公网安备 11010802042870号
京ICP备2023020795号-1
分享
求助内容：
应助结果提醒方式：
扫码关注我们
