12:00 PM - 12:15 PM
[HDS11-11] Application of AI Image Recognition in Rip Current Education
Keywords:Rip Currents, Artificial Intelligence, Image Recognition, Disaster Education
Globally, approximately 80% of beach accidents each year are caused by rip currents (Rip Currents—Summer Beach Killers, National Taiwan Ocean University Teacher Training Center, Associate Professor Cheng-Chieh Chang, United Daily News Weekly, 2017.1.23), making them a major source of risk in coastal areas. With the impact of climate change, the frequency and intensity of rip currents are expected to increase, posing growing threats to coastal communities and tourists. However, traditional disaster education methods are limited in immediacy and effectiveness, making it difficult to address rapidly changing risks in dynamic environments.
This study aims to develop a mobile application based on AI real-time image recognition technology, integrating efficient recognition, risk alerts, and educational functions. The application focuses on early warning of rip currents and disaster education. By employing AI technology trained with deep learning models, the system optimizes recognition performance for diverse sea conditions and is tested in academic institutions to evaluate its impact on users' risk perception and self-rescue capabilities.
The anticipated results indicate a significant improvement of over 30% in users' accuracy in risk recognition. Specifically, in environments with varying lighting and sea conditions, the app provides real-time feedback, enhancing educational outcomes. This research demonstrates the potential of AI technology in natural disaster education, offering an innovative solution for addressing rip current disasters and showcasing its scalability to other similar disaster scenarios.
This study aims to develop a mobile application based on AI real-time image recognition technology, integrating efficient recognition, risk alerts, and educational functions. The application focuses on early warning of rip currents and disaster education. By employing AI technology trained with deep learning models, the system optimizes recognition performance for diverse sea conditions and is tested in academic institutions to evaluate its impact on users' risk perception and self-rescue capabilities.
The anticipated results indicate a significant improvement of over 30% in users' accuracy in risk recognition. Specifically, in environments with varying lighting and sea conditions, the app provides real-time feedback, enhancing educational outcomes. This research demonstrates the potential of AI technology in natural disaster education, offering an innovative solution for addressing rip current disasters and showcasing its scalability to other similar disaster scenarios.