Method and system for forecasting crowd dynamics

• Main Authors: KUROSE MAI, YONETANI RYU
• Format: Patent
• Language: Chinese; English
• Published: 17.06.2022
• Subjects: CALCULATING; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

A method for modeling crowd movement, comprising: obtaining a time series of images of a physical venue; and for each of the images, subdividing the respective image into respective sets of logical pixels according to a predetermined mapping. For each logical pixel of each image, the method calculates a respective crowd density representing a respective number of moving objects per unit area in the physical venue at the logical pixel, thereby forming a time series of crowd density maps corresponding to the time series of images. The method then uses the continuous crowd density map pairs to train a model regarding spatio-temporal variations in crowd density at the physical venue. A method of predicting a future crowd density map at a physical venue using a current image of the physical venue and the trained model is also disclosed. 一种用于对人群运动进行建模的方法，包括：获得物理场馆的图像的时间序列；以及针对图像的每一个，根据预定映射将相应图像细分成相应逻辑像素集。对于每个图像的每个逻辑像素，该方法计算相应的人群密度，该人群密度表示在逻辑像素处的物理场馆中每单位面积的移动对象的相应数量，从而形成对应于图像的时间序列的人群密度图的时间序列。然后，该方法使用连续的人群密度图对来训练关于物理