Old people tumble detection method based on neural network

Abstract

The invention discloses an old people falling detection method based on a neural network, and the method comprises the following steps: S1, carrying out the preprocessing of depth image data, and obtaining a binary image; s2, obtaining an optimal ELM mapping parameter by adopting a particle swarm optimization algorithm to obtain an optimized ELM neural network model; s3, carrying out the sample processing of the binary image, adding a label to a sample, inputting the sample into an optimized ELM neural network model for training, and obtaining a feature vector; s4, inputting the feature vectors into a convolutional neural network model to obtain the probability that the old people are in a falling state; and S5, according to the probability that the old people are in the falling state, judging whether the old people have a falling behavior or not. The problems that the number of consumed neurons is too large and the stability of a feature extraction result is poor due to random generation of an input layer weight and an implicit layer deviation in falling detection based on traditional machine learning are solved.

Description

technical field [0001] The invention belongs to the technical field of human body recognition, and in particular relates to a fall detection method for the elderly based on a neural network. Background technique [0002] At present, my country's aging process is accelerating. With age, body function and balance control decline, leading to an increased risk of falls. Falls in the elderly can easily cause physical injury and affect daily life. Fall detection is the basis for implementing fall protection and provides support for fall injury assessment and timely rescue. Fall behavior detection mainly senses human behavior data through sensors, preprocesses the data and extracts features, and finally performs fall recognition. Sensors for fall detection can be classified into wearable sensors and environment awareness sensors. Wearable sensors need to be worn with you and are suitable for daytime outdoor locations. Environmental sensing sensors include video, infrared and s...

AI Technical Summary

Problems solved by technology

[0006] Aiming at the above-mentioned deficiencies in the prior art, a neural network-based fall detection method for the elderly provided by the present invention solves the fall detection based on traditional machine learning, and the input layer weights and hidden layer deviations are randomly generated, which will cause consumption The number of neurons is too large, and the stability of the feature extraction results is poor

Images