Bio-Info-Sensor Image Processing Approach: Disaster Pre-alarm for Earthquake

Abstract

Earthquake is one of the major problems and there is no perfect & accurate solution for this. Even we can not prevent natural earthquake from occurring and must face terrible effect and major loss due to its result. In this paper, we are going to present a concept and an expected architecture which would detect or tend to detect earthquake prior to its occurrence. It will use bio-science and information technology to develop the whole system. This paper gives the blueprint of how the system will look like after merging the two major entities of physical world? Bio-science provides us the senses/sensors while the information technology provides the measures of complex computing/processing related to large amount of data or calculations automatically with high accuracy and speed. In this case, animals behave as precursors. But despite lots of research, today researchers are not capable of analyzing abnormal behavior of animals under various circumstances because it is very complex and highly computational task. Therefore, we are proposing an idea of making use of “Automatic Behavior Recognition System,” which uses computer vision technology. In this system, the signals are captured via static cameras and the recordings would be processed by the behavior analysis algorithm which is further linked with a central database server having detailed information about the various behavioral activities under diverse circumstances of several animals, finally provides a high-level textual report and decision regarding the animal’s abnormal behavior for advance researches and the betterment of the system.