The present invention provides a method and
system of monitoring the aged through sound by means of an
internet of things. The voice recognition care
nursing system comprises a host, at least one submachine and a care
nursing unit. The submachine is installed in a room and configured to detect plural sound signals existing in the environment at the same time, and it is determined whether the sound signals are voice signals or non-voice signals or not through determining whether the sound signals have
baseband-
overtone structure or not; if the sound signals are voice signals, the content of the words and expressions of the voice signals are identified, and it is determined whether the sound signals include one of preset plural help information or not; if the sound signals are non-voice signals, it is determined whether the sound signals are one of preset plural abnormal sounds or not; if the content of the words and expressions of the voice signals is help information or the non-voice signals are abnormal sounds, the host is configured to send warning information to the care
nursing unit and inform family members or a caregiver of inspection.