Development of childhood fall motion database and browser based on behavior measurements.

Highlights: [•] We developed a sensor home equipped with an embedded video-surveillance system. [•] We developed a fall detection algorithm and a body axis tracking algorithm. [•] We measured behaviors of 19 children aged 11–50 months old at the sensor home. [•] Fall database includes 105 incidents, and its browser enables conditional searches. [•] The database enables accurate estimation of injury risk resulting from fall. [ABSTRACT FROM AUTHOR]