A framework of ontology-based knowledge information processing for change detection in remote sensing data
Shutaro Hashimoto, Takeo Tadono, Masahiko ONOSATO, Masahiro Hori, Takashi Moriyama · 2011
This paper proposes a new framework for automatic interpretation, especially for change interpretation, in order to implement insightful and versatile interpretation like manual interpretation. The framework considers remote sensing data analysis as a knowledge information processing. For handling large amount of knowledge on computers consistently, the framework employs double-layered knowledge structure; ontology and heuristic. The ontology is a conceptual dictionary about remote sensing domain, and heuristic is experimental knowledge containing uncertainty. Additionally, this paper introduces the method for constructing Bayesian network using knowledge, and inferring target objects and changes using it. An experiment of change detection is conducted to demonstrate our framework.