Beseye Cloud Security has provided a cloud-based surveillance software solution for Japan's Tokyu, an operator of commuter railroad lines in the Tokyo metropolitan area, to prevent purposeful or careless intrusion into railroad crossings.
While most surveillance solutions using AI-based image analysis rely on facial recognition, Beseye said it specializes in skeleton-based action recognition to analyze human dynamic behavior. The skeleton-based action recognition first detects people entering or lingering in a certain area from images captured by surveillance cameras, and then identifies the persons from their faces and detects what they are doing there to analyze their behavior, Beseye explained.
Tokyu has adopted Beseye-developed software solution for all surveillance systems installed at railroad crossings, and when the surveillance systems detect persons intruding into railroad crossings or railroad tracks, they will automatically notify control centers or responsible staff members to prevent the persons from being hit by trains, Beseye indicated.
Surveillance software solutions can be customized to cater to various purposes such as at production lines, retail stores, hospitals, transportation stations and homes, Beseye said. Such solutions can be paired with surveillance cameras developed by Beseye or existing equipment already installed by clients, Beseye noted. Simply speaking, such solutions add software-based functions to make existing surveillance hardware smart and automatic, Beseye indicated.