P‐3.6: A training system for the rescue process of acute and severe diseases based on AR

Abstract

China is one of the countries with the most serious natural disasters in the world. Hidden dangers and risks of all kinds of accidents are intertwined and superimposed, and the factors affecting public safety are increasing. It is urgent to strengthen the construction of emergency medical rescue capacity. During the nineteenth collective learning of the Central Political Bureau, general secretary Xi Jinping stressed that we should strengthen the capacity building of aviation emergency rescue and improve the airspace guarantee mechanism for emergency rescue. As the most important aerial medical evacuation platform, helicopter has the advantages of fast response, strong mobility, wide rescue range, direct arrival at the first line of treatment, etc. it has become an important means of modern rescue, and also the only way to open life treatment in case of land damage. From the international experience, the helicopter emergency medical system established by most developed countries plays an important role in disaster relief and anti‐terrorism. China has a large population, complex terrain and uneven regional development. In case of disasters, the casualties are often more severe and the demand for treatment is more urgent. It is urgent to establish an aviation medical emergency rescue system suitable for the national conditions and with helicopters as the main carrier, and accelerate the research and development of key technical equipment.In recent years, more and more people realize the necessity and benefits of the application of interactive technology for aviation medical simulation training in major events and disasters. Interaction refers to the transmission and exchange of intelligence, data, data and technical knowledge in all aspects of nature and society. From the perspective of information theory, local chronicles are a tangible text information carrier, which collects all kinds of information in a certain region. Interactive technology is to use certain means to achieve the purpose of interaction, and gradually enter the era of multi field application! In thefield of aviation medical rescue, it is in urgent need of penetration and integration.This paper mainly develops an AR emergency rescue process training system. Through the creation of data sets, action recognition based on computer vision and deep learning, the selection of recognition model and action capture, the action is recognized based on AR, and then 22 joint points of human hand and upper arm skeleton are taken to label the data space position. Through the above learning results, we can get the real‐time hand recognition based on depth information Upper arm skeleton output. Neural network is used to recognize human skeleton action. The skeleton information coordinates are imported into RNN neural network for deep learning. Finally, based on the complementarity of CNN and LSTM, combined with the idea of dual stream structure network, the action recognition is carried out.