Research On Multi-user Human-computer Interaction Technology In Simulated Shooting Training System

With the gradual strengthening of the informatization construction of the troops,higher and higher requirements have been placed on the quality of soldiers' training.Shooting training,as the basic subject of military training,is of great significance to trainers to improve their military skills.At present,in the military shooting training,mainly the live-fire shooting training,not only the high risk coefficient,the training cost is high,but also the environmental requirements are more demanding,and the training method cannot be flexibly changed according to the training demand.The method of automatically counting the shooting results by using the traditional image processing technology is easily affected by the ambient light,and it is difficult to accurately identify the target on the basis of ensuring flexibility.In this thesis,the application of computer vision technology in simulation training system is studied.Firstly,the camera calibration algorithm is studied and its formula is deduced.Then,based on the existing image processing based target detection method,two kinds of spot detection and localization algorithms are proposed and compared experimentally.The simulated training system is composed of training user terminal,background processing system and terminal processing system.The spot data is collected by the camera in real time.The collected video stream is analyzed and processed by the OpenCV and PyQt image interface development framework on the computer.Spot position,on the one hand,using the Unity3D game development engine to build a simulation shooting training system application scenario,combined with the Arduino joystick control physical object,simulation of the virtual aircraft in the virtual reality scene,simulation of the actual battlefield training environment,increase simulation training Realism.On the other hand,the design realizes that the back-end system can summarize,analyze and visualize the data of the shooting training of multiple users,and manage the training data of multiple users.The experimental results show that the designed system can achieve 96%recognition rate in the indoor projection screen.The average horizontal and vertical error of the impact point detection are).013 and 0.021 respectively,which can achieve the basic goal of indoor shooting training and reduce the training cost.Reduce the difficulty of operation and improve the training effect.Shooting training for most simple needs,on the basis of ensuring a certain detection accuracy,can be easily deployed and installed.