UAV画像における植生観測のためのカゲ補正手法

抄録

Phenology observation based on biological reactions is one of the methods for predicting impacts on climate change. A variety of observation platforms are used for phenology observations, but among them, observation by UAV with hyper spectrum resolution, hyper spatial resolution and hyper temporal are expected. However, the results of observations by UAV are affected by change of light sources. Therefore, this study focuses on the characteristics of the light source during the observation and corrects it. This method used 47 scenes observed between June 2019 and June 2020.Due to the different types of light sources, the observed scenes on sunny and cloudy days were corrected using different procedures. The scattered light on a cloudy day was corrected by a sky view factor. On a sunny day, the light source was masked by shadow simulation using Digital Surface Model and separated into direct light and scattered light. The direct light was applied topography correction and the scattered light was applied sky view factor correction. As a result, both sunny and cloudy days were corrected with an accuracy of ±5 digital count value in Jpeg images for the corrected brightness value. Finally, this method allowed to see the phenology of the vegetation throughout the year.