Intelligent Fall Detection System based on Sensor and Image data for Elderly Monitoring

In recent years Internet of Things has gained much attention. Various techniques are being developed to acquire the data from human environment for various smart services and applications. Monitoring elderly people remotely using smart homes is one of the difficult and challenging tasks in lOT due to many accidents that occur during daily activities like falls etc. In elderly people, falling is considered as a main reason for cause of injuries of death. To prevent the fall of elderly people and to identify the fall in smart homes technology need to be increased so that the fall can be prevented and rate of survival can be increased. Currently many loT, wearable devices, mobile phones etc have been developed with advanced technologies which can be used to develop a fall detection technique using loT and Deep Learning. In this paper an loT and Deep learning based elderly fall detection technique is proposed. Deep convolutional Network is used to extract the features obtained using loT devices and to detect the fall using intelligent Deep fall Detection algorithm. Feature extraction is performed using Squeeze Net model by extracting required features. SVM classifier is used for classification of fall and non -fall occurrences. If a fall event is detected an alert message is sent to the smartphone of hospital management. UR fall detection dataset is utilized to perform the experiments. Based on the obtained results it is inferred that proposed method achieved an accuracy of 99.81 % on fall detection dataset.