JAPAN: Following successful trials, Tokyo Metro has decided to roll-out the use of stereo depth cameras to provide real-time information on train occupancy.
The system developed with Ueno Green Solutions and Cyber Core uses cameras with two closely-spaced sensors a known distance apart, and software which compares the images from each sensor to estimate depth and generate a real-time measurement of crowding within the train.
In the past Tokyo Metro has estimated crowding on trains using vehicle weights and data from ticket barriers, but this was complicated by the number of operators sharing tracks.
The first trials with the cameras were undertaken in September 2019. Further trials were spurred by the need for higher quality information during the Covid-19 pandemic, and a second experiment was started in November 2020.
