期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2010
卷号：XXXVIII - Part 1
出版社：Copernicus Publications
摘要：Recently, activity-based analysis using GPS equipment as data collector has being a hot issue. But most researches focus on data from wearable GPS recorder for person because of easy detailed activity logging and interactive validation with users. Nevertheless, available floating car data and geographic context databases provide the possibility for activity-based analysis based on GPS vehicle tracking trajectories, this paper presents a novel and efficient approach to automatically identify activity-locations as well as activity- types from raw GPS vehicle tracking data. A number of contextual variants related to the activity locations including temporal information, spatial information and probability information are considered with the help of digital map containing P oints of Interest (POIS). Taking the three aspects of inputs mentioned above into a Multi-Variants Analysis framework, detailed information of each activity will be identified. Finally, experiments using real world floating car data are conducted for evaluation and to show how this approach will help in varieties of applications in both traffic and activity-based analysis
关键词：Activity Identification; GPS Trajectories; Multi-Variants Analysis; Intelligent Transportation System
