Visualizing the dose rate distribution around the Fukushima Daiichi Nuclear Power Plant using artificial neural networks

Sasaki, Miyuki  

This study proposes a method of visualizing the ambient dose rate distribution using artificial neural networks (ANN) from airborne radiation monitoring results. The ANN method was applied to the results of the airborne radiation monitoring which was conducted around the Fukushima Daiichi Nuclear Power Plant by an unmanned aerial vehicle. The ANN was constructed by training data consisting of input variable dataset (radiation count rate, altitude, topographic data, photographic RGB data) and objective variable dataset (the air dose rate data at 1 m above the ground level). The reliability of the ANN method was evaluated by comparison with the ground-based survey data. The dose rate map created by the ANN method reproduced ground-based survey results better than traditional methods.