LEVERAGING VITAL SIGNS CLASSIFICATION FOR DISASTER MANAGEMENT AND ENVIRONMENT RISK ASSESSMENT TO SAFEGUARD ECOSYSTEM

Abstract

Monitoring the earth vital signs is need to check the earth Condition and whether to ensure the safety of the living beings The existing system employs the Multinomial Naïve Bayes algorithm for image-based detection of Earth's vital signs, including seismic activities, cyclones, floods, and wildfires. While effective in its simplicity, the algorithm's assumption of feature independence may limit its ability to capture intricate relationships within image datasets. These systems may struggle to adapt to the dynamic nature of environmental. In the context of Earth monitoring, the Naive Bayes model is employed to analyse data related to seismic activities, cyclones, floods, and wildfires. However, the inherent assumption of independence among features in the Naive Bayes model may lead to limitations in accurately capturing complex relationships within the diverse and dynamic datasets associated with Earth's vital signs. The proposed system is CNN with the VGG16 model and Random Forest algorithm (Ensemble Learning model). The VGG model based on CNN architecture it is used to extract the features from the input image, preprocessing and train and test the data, then the Random Forest algorithm is used to the predict accuracy and labels for input data. The advantages of earth vital signs is real-time monitoring, and predictive analytics and improve accuracy. Versatile tool for environmental surveillance and enabling the development of early warning systems for prompt responses to environmental threats and natural disasters.