Development of a Detection System for People Drowning Through Aerial Images and Convolutional Neural Networks

Ramos J.O.; Flores H.V.; Villena J.R.; Gonzales J.Z.; Joseph G.M. · Przeglad Elektrotechniczny (2023) · DOI: 10.15199/48.2023.02.18

Resumen

This work develops a system capable of determining the presence of a person in the water by classifying a convolutional neural network (CNN). When a drowning is alerted, the drone camera takes aerial captures of the water area, the first trained CNN is in charge of determining if it is deep water or not, while the second CNN is in charge of identifying the presence of a person in that video frame; in case both detections are positive, the drone will drop the life ring it carries. The tests confirm that this system is capable of providing a means of survival to the person in an ideal time.

Instituciones: Universidad Nacional de San Agustín de Arequipa