Research On Simulation Method Of Group Evacuation Behavior Driven By Video

In recent years,with the development of social economy and the acceleration of urbanization,the population density of our country has increased rapidly,and various large-scale public buildings are also increasing day by day.The crowds in cities are highly concentrated,especially in public places such as schools,subway stations and railway stations.The crowding phenomenon is especially obvious during the peak period of holidays and spring transportation.In general,the design of large-scale buildings is relatively complex.In case of explosion,earthquake,fire and other emergencies,how to quickly evacuate all the personnel inside the building to the outside of the building is an urgent problem.Research shows that in case of emergency such as fire and earthquake,people are extremely panicked,unable to find a safe exit quickly and reasonably,and trampling accidents are prone to occur in high crowd density scenarios,which reduces evacuation efficiency and causes loss of life and property.Because of the economic and social factors,it is almost impossible to use real crowd drills to prevent,so the crowd evacuation can be simulated by computer simulation technology when an emergency occurs.At present,although the crowd evacuation simulation model can realize the evacuation process of group movement,its simulation effect still has a certain gap with the real pedestrian movement.On the one hand,the group movement path planned by the simulation model is not exactly the same as the escape path of the crowd in the real scene.On the other hand,the existing simulation models cannot realistically restore the pedestrian avoidance effect in the real scene.To solve the above problems,this paper proposes a video-driven crowd evacuation behavior simulation method.In this method,building scene features are extracted from real video,safety evacuation landmarks are set as crowd evacuation guiding points,and direction parameters are set for these points according to the direction of emergency exit.The optimal evacuation path is finally obtained by ESPRM algorithm and stored in the path graph.The parameters of crowd movement are extracted from video data,and these parameters are applied to RVO obstacle avoidance model.To achieve the avoidance behavior of pedestrians in the process of simulation,so as to achieve a more realistic simulation effect.Finally,a video-driven crowd evacuation simulation system is constructed to simulate and validate the scene of teaching building and large outdoor scene.The main work and innovations of this paper are as follows:1.A new global path planning algorithm based on secure evacuation signs--ESPRM algorithm is proposed.Firstly,the method analyzes the role of the safety evacuation signs in the global path planning,and extracts and analyzes the building features and safety evacuation signs in the real video,and obtains the key points in the path planning as the evacuation flag information applied to the PRM algorithm.Finally,the evacuation path obtained by the algorithm is closer to the real situation,which makes the simulation result have higher practicability.2.A video data driven RVO model is proposed.For the RVO model,it is assumed that all individuals change the movement speed in addition to the trigger avoidance strategy conditionduring the whole evacuation process.Under normal circumstances,the motion is uniform.The situation of the descendant's movement speed is affected by many aspects.The influence of the degree of crowding in the video on the speed of pedestrian movement abstractly quantifies the negative correlation between the speed of movement and the density of pedestrians and applies it to the avoidance strategy.At the same time,through the abstraction and analysis of the crowd movement data in different states of the real video at different times,the Pathfinder tracking software is used to extract the individual coordinates of the pedestrian,the number of individuals,the individual instantaneous movement speed and other parameters from the real video,and apply these parameters to In the evacuation simulation,the initialization of different pedestrians and the diversity of different individual movement speeds at the beginning of the simulation experiment are made.Since the above data are derived from real video,the authenticity of the evacuation simulation is guaranteed,and the speed change of the pedestrian in the evacuation simulation process is more realistic.3.Construct a video-driven group evacuation simulation system.The system is developed based on Microsoft Visual Studio 2013 and MFC,and is applied to the realistic rendering platform to implement functions such as scene modeling,global path planning,group motion simulation and motion state display.Taking the teaching building scene and the scene of Jinan Quancheng Square as an example,the evacuation simulation research and analysis are carried out,and the simulation results are displayed using the realistic rendering platform rendering output.The experimental results show that the proposed method can simulate the group evacuation process more realistically and has certain reference value and practical value.