Prevention of children falls from balconies using RGB-D cameras and a digital human child model
Hiroki Yamamoto, Koji Kitamura, Yoshifumi Nishida, Hiroshi Mizoguchi · 2017
Based on the realization that more than 100 Tokyo children have died after falling from balconies during the last five years, this paper proposes a new system for detecting potentially dangerous situations using red green blue-depth (RGB-D) cameras, and thus aims to help prevent children from falling from balconies. The system, which is based on the results of our investigation into how residents use house and apartment balconies, has two functions. The first is to detect a solitary child on a target balcony by estimating his or her age from the observed body trunk length. The second is to detect places that the child could climb by evaluating the shapes of objects on the balcony. The effectiveness of these functions was confirmed via experiments at a mockup living laboratory and a real-world living laboratory installed in an ordinary high-rise apartment.