In this approach, the ground is assumed to have a flat form, and this is called the flat source model (FSM) 7. In a conventional approach, the count rates and the pulse height distributions collected by the UAV-survey are converted into air dose rates and deposition of radioactive cesium on the ground.
![iaea tecdoc 1363 iaea tecdoc 1363](https://i1.rgstatic.net/publication/314273323_Processing_of_Airborne_Gamma-Ray_Spectrometry_Using_Inversions/links/5a01f817a6fdcc55a158d9b0/largepreview.png)
However, a UAV-survey is less accurate than a ground-based radiation survey using a handheld survey meter because it has a long distance from the ground surface source. UAVs are effective tools for data collection over wide areas that are located around FDNPP because a person does not have to approach a dangerous place 5, 6. As a quick and efficient survey method, manned helicopters and unmanned aerial vehicles (UAVs) have been developed to visualize the environmental distribution of the air dose rate in the airborne radiation survey (UAV-survey) after the FDNPP accident 3, 4. Nine years after the accident, the ambient dose rate (air dose rate) has been decreased by radioactive decay, decontamination work, and weathering effect within the 80-km radius zones from the FDNPP 2.
![iaea tecdoc 1363 iaea tecdoc 1363](https://ars.els-cdn.com/content/image/1-s2.0-S096980431730859X-gr5.jpg)
Large quantities of radionuclides were released into the atmosphere after the Fukushima Daiichi Nuclear Power Plant (FDNPP) accident in March 2011 1. The dose rate map created by the ANNs method reproduced ground-based survey results better than traditional methods. The reliability of the ANN method was evaluated by comparison with the ground-based survey data. This number of training case was considered a rough criterion of the required number of training cases. The quantitative evaluation index (the root-mean-square error) between the ANN-converted value and the ground-based survey result converged at 200 training cases. The number of training cases was related to the error between the ground and converted values by the ANN.
![iaea tecdoc 1363 iaea tecdoc 1363](https://www.researchgate.net/profile/Jesus-Cortes-Rodicio/publication/236009660/figure/fig4/AS:393461313359875@1470819895294/Resultados-para-Superposicion-Los-puntos-indican-la-discrepancia-entre-calculo-y-medida_Q320.jpg)
Much of the survey data obtained in the past were used as the training data for building a network. The 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. This study proposes a new method of visualizing the ambient dose rate distribution using artificial neural networks (ANNs) from airborne radiation monitoring results.