摘要
通过改进传统的社会力模型,建立了一种适用于密集场景中个体目标方向预测的目标预测模型.该模型详细定义了个体在人群中受到的驱动力、躲避力和排斥力,以及3种力的合力对个体运动状态的影响,并由此实现对个体运动目标方向的预测.将该模型应用于大量密集场景视频进行实验测试.结果表明,该模型能够对个体目标方向进行正确估计;同时,该模型能够为更多密集群体行为的相关应用提供研究基础.
A goal estimation model is built based on the traditional social force model.It is used to predict the directions of individuals’motivations in the crowd.The model particularly gives the definitions of the driving forces,evasive forces and repulsive forces exerted on the individuals.The prediction of the goal is based on these three forces.A lot of experiments on crowd scenes demonstrate the effectiveness of the model.Meanwhile,the model is practical for plenty of crowd related research topics.
出处
《上海电力学院学报》
CAS
2018年第1期81-84,共4页
Journal of Shanghai University of Electric Power
基金
国家自然科学基金(61401268)
上海市自然科学基金(15ZR1418400)