An approach for automatic change inference in high resolution satellite images

Recently, more detailed change detection is becoming possible due to increases in the availability of satellite imageries for practical use and improvements in their spatial resolution. While the amount of data is increasing, manual interpretation is still being used as a conventional method of change detection. For these reasons, practical change detection techniques are required. This paper proposes a knowledge-based change detection approach, which can obtain change information that includes not only land cover changes, but also contextual changes, such as types of damage caused by natural hazards. This approach mainly consists of two processes: information extraction and change inference using Bayesian network. Information extraction employs object-based image analysis for extracting spatial information. Change inference uses extracted information and the Bayesian network constructed from knowledge of change detection process. To demonstrate this approach, change detection of mudslide damage caused by heavy rain in Yamaguchi Pref., Japan was conducted.