Volume XLII-2/W7
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-2/W7, 1481-1486, 2017
https://doi.org/10.5194/isprs-archives-XLII-2-W7-1481-2017
© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 4.0 License.
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-2/W7, 1481-1486, 2017
https://doi.org/10.5194/isprs-archives-XLII-2-W7-1481-2017
© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 4.0 License.

  14 Sep 2017

14 Sep 2017

LABELING RESIDENTIAL COMMUNITY CHARACTERISTICS FROM COLLECTIVE ACTIVITY PATTERNS USING TAXI TRIP DATA

Y. Zhou1,3 and Z. Fang2 Y. Zhou and Z. Fang
  • 1Wuhan Land Use and Urban Spatial Planning Research Center, 55 Sanyang Road, Wuhan, China
  • 2State Key Laboratory for Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan, China
  • 3School of Remote Sensing and Information Engineering, Wuhan University, Wuhan, China

Keywords: Taxi trip data, activity space, residential community, land use classification, mobility patterns, O/D allocation

Abstract. There existing a significant social and spatial differentiation in the residential communities in urban city. People live in different places have different socioeconomic background, resulting in various geographically activity patterns. This paper aims to label the characteristics of residential communities in a city using collective activity patterns derived from taxi trip data. Specifically, we first present a method to allocate the O/D (Origin/Destination) points of taxi trips to the land use parcels where the activities taken place in. Then several indices are employed to describe the collective activity patterns, including both activity intensity, travel distance, travel time, and activity space of residents by taking account of the geographical distribution of all O/Ds of the taxi trip related to that residential community. Followed by that, an agglomerative hierarchical clustering algorithm is introduced to cluster the residential communities with similar activity patterns. In the case study of Wuhan, the residential communities are clearly divided into eight clusters, which could be labelled as ordinary communities, privileged communities, old isolated communities, suburban communities, and so on. In this paper, we provide a new perspective to label the land use under same type from people’s mobility patterns with the support of big trajectory data.