Sistema de detección de caídas en personas utilizando vídeo vigilancia.

Falls are the second cause of death in the world. This is the reason why systems of falls detection, framed in the telemedicine field, are widely investigated nowadays. However, these systems often result in false positives. In this paper we propose the development of a fall detection system for people. The system is based on the use of a Kinect 2.0 sensor that allows monitoring from one to six patients and an alarm system that sends both a message and an image to the e-mail of the person who is taking care of the patient. The biomechanical analysis of the fall, carried out using iPi Mocap Studio program, and the fall detection algorithm, developed in LabVIEW graphic programming software, allowed measuring the position and velocity values of a selected point in the skeletonization of the person. Different actions were simulated, including walking, sitting, bending and falling. From the position of the point-selected point (spinebase) of the skeleton of the person, the velocity was calculated with the derivative. This characteristic allowed defining the linguistic variables and fuzzy sets to classify the type of movement. The system was validated with 3 people, of different ages and heights, in a controlled environment, achieving an effectiveness of 95.8% and the activation of the alarm system, creating a technological assistant of adequate and reliable aid that guarantees the integrity of the patient. [ABSTRACT FROM AUTHOR]