Abstract:
It is significant for geological disasters detection from remote sensing image in emergency rescue. However, the automatic detection methods for geological disasters, dep...Show MoreMetadata
Abstract:
It is significant for geological disasters detection from remote sensing image in emergency rescue. However, the automatic detection methods for geological disasters, depending only on low-level imagery features, generally result in low recognition precision. In practice, the most current extraction approaches of geological disasters are manual visual interpretation with the aid of experts' knowledge. The means have been inefficient and expensive. Accordingly, the aim of this paper is improving the recognition accuracy and efficiency of geological disasters. On the basis of object-oriented image segmentation, the study proposes a method for geological disasters detection with experts' knowledge and image features comprehensively. The results show that this method is more accurate than pixel-based classification and object-oriented image classification only with image features and more efficiency than manual visual interpretation. The study will promote the researches of disaster information extraction from remote sensing images automatically and intelligently.
Date of Conference: 10-15 July 2016
Date Added to IEEE Xplore: 03 November 2016
ISBN Information:
Electronic ISSN: 2153-7003