Technology of NEC provides for congestion and flow of crowds … – Tech&Net

THE NEC Corporation has announced the strengthening of its technology analysis of the behavior of crowds (* 1), allowing to estimate and predict with accuracy and in real time, the degree of congestion of a crowd and the flow of people, using specialized systems of security cameras.

This newly developed technology, which belongs to the innovative portfolio of technologies of artificial intelligence "NEC the WISE" (*2), parses the images from the security cameras to create data that define with precision the degree of congestion and the flow of people in the environments (* 3) consisting of tens of thousands of people, including large facilities or venues of events. In addition, it employs the model of the crowd the owner of the NEC, to simulate the behavior of people in an environment overcrowded, simulating the way people follow or avoid collisions among themselves, to optimize the analysis and to estimate or accurately predict the various situations in real-time.

"This technology predicted the degree of congestion in areas of a major sporting event 10 minutes in advance with a margin of error of 20%. This allows the orientation of crowds in real time, helping to avoid congestion and to increase safety", said Akio Yamada, General manager, Data Science Research Laboratories, NEC Corporation.

1 – Understanding quantitative degree of congestion and the flow of people, even in an environment overcrowded by analysis of footage from security.
When you combine technology that allows you to view the conditions of a crowd from security images (the degree of congestion and the estimated number of people with a technology to extract the motion existing in these images, the NEC has developed a new technology to estimate multiple elements of the movement of a crowd, allowing a quantitative understanding of the number of people moving in each direction.

This technology allows you to accurately estimate the degree of congestion and the distribution of the flow of a crowd within the field of view of a camera, integrating and analyzing the movements of the crowd. As a result, it is possible to understand the number of people that pass by a certain area in each direction, even in an environment that is very overcrowded, where people apparently tend to overlap each other.

2 – the Forecast of the conditions of a crowd for up to ten minutes in advance based on information about the degree of congestion and the flow of people obtained from several places. NEC has developed a model of crowd to estimate and predict the flow of people in real time, by incorporating information about the density and the flow of a crowd, for later analysis.

The model of the crowd expressed the behavior of people in an environment that is overcrowded and displays the movements of each agent, such as individuals who move to avoid collisions and to individuals that travel through empty spaces, with the goal of accurately simulate the conditions of the multitude.

*1: http://www.nec.com/en/press/201311/global_20131107_02.html

*2: http://www.nec.com/en/press/201607/global_20160719_01.html

Studies NEC on Artificial Intelligence

http://www.nec.com/en/global/rd/crl/ai/index.html

*3 Environment of the crowd: an environment in which there are more than two people per m2 in terms of available space for walking, and four people per m2 in terms of space available for people stops.

Via press release