An Environmental-Semantic Computing System Of Multispectral Imagery For Coral Health Monitoring And Analysis

The global environmental analysis system is a new platform to analyze environmental multimedia data that acquired from nature resources. This study aims to realize and interpret coral reefs phenomena and changes occurring that happening in global scale by utilizing Acropora coral as a bioindicator or natural sensing. This paper presents a new environmental-semantic computing system of multispectral imagery for automated coral health monitoring and analysis to realize and recognize coral condition in actual situation. Multispectral semantic-image space for coral monitoring and analysis can be utilized for ocean environment monitoring and assessment by measuring coral reef health which highly beneficial to the current ocean pollution problem. Our method applies semantic distance calculation to measure similarity between multispectral image data and context images including three coral conditions (healthy, bleaching, and dead). In our experiments, we applied the SPA function which is an effective concept to design environmental systems with Physical-Cyber integration. This paper presents case study of Acropora coral monitoring and assessment at Man-nai Island, Rayong province, Thailand. Therefore, an additional objective of this research is to apply the Artificial Intelligence (AI) and Environmental monitoring system for combatting the ocean pollution problem by transferring the knowledges and technology from computer science fields to fundamental marine science research.