Volume IV-3/W1
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., IV-3/W1, 65-70, 2019
https://doi.org/10.5194/isprs-annals-IV-3-W1-65-2019
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., IV-3/W1, 65-70, 2019
https://doi.org/10.5194/isprs-annals-IV-3-W1-65-2019
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

  01 Mar 2019

01 Mar 2019

DEVELOPMENT OF METHODOLOGY FOR PLANT PHENOLOGY MONITORING BY GROUND-BASED OBSERVATION USING DIGITAL CAMERA

M. Yamashita1, Y. Shinomiya2, and M. Yoshimura3 M. Yamashita et al.
  • 1Faculty of Agriculture, Tokyo University of Agriculture and Technology, Fuchu, Tokyo, Japan
  • 2Pasco Corporation, Meguro-ku, Tokyo, Japan
  • 3Center for Spatial Information Science, The University of Tokyo, Meguro-ku, Tokyo, Japan

Keywords: Digital camera, Phenology, Long-term observation, Image indices

Abstract. When monitoring phenology at ground level, it would be more important to continue observations in long terms and to detect the timing of various phenological events such as leafing, flowering and autumn senescence. In this study, to develop the methodology for plant phenology monitoring by using digital camera, we examined how multiple image indices, which are derived from multi-temporal visible images, respond to the changes of colors of leaves and flowers for several target species of plants, and tried to detect various phenology events by tracing time series changes of the coordinate in the feature spaces of two indices. As a result, we found out that it was possible to understand the characteristics of the phenological events for different species from each image index. Also, it was identified that the utility of combination with two indices would be effective to detect the timing of phenology events in the feature space of two indices. In the actual phenology monitoring, it would be effective to use a single index for understanding the seasonal characteristics and to use the combination of two indices for detection of the timing of phenology events by tracing the time series changes in the feature space.