Case studies of automatic change detection using AVNIR-2 onboard ALOS

This paper suggests a new automatic change detection approach using image-object-based contextual inference. The approach first detects changed areas, and then infers what happened there. The inference process uses knowledge based on a change detection process performed by humans. It enables flexible and intuitive description of the change detection process. In this paper, we describe the results of case studies on automatic change detection using the Advanced Visible and Near Infrared Radiometer type-2 (AVNIR-2) onboard the Advanced Land Observing Satellite (ALOS, nicknamed “Daichi”).