Safety monitoring system based on multi-sensor

Abstract

The present disclosure relates to a multi-sensor based safety monitoring system. The system includes: a sensor device end, used to collect the environmental information of the target user, and send the environmental information to a local server; a local server, used to input the environmental information as a training sample into the artificial neural network model for automatic Learning and training to update the parameters of the artificial neural network model; the cloud server is used to send the updated parameters to the sensor device end, and receive the security monitoring of the target user returned from the sensor device end according to the updated parameters status, and push the security monitoring status of the target user to the client; the client is configured to receive the security monitoring status of the target user. The disclosure uses multi-dimensional sensing technology and self-learning algorithms based on big data to learn and understand the living habits of target users such as the elderly, so as to achieve the ability to predict potential risks.

Description

Technical field [0001] The present disclosure relates to the field of artificial intelligence, in particular to a multi-sensor-based security monitoring system in the Internet of Things. Background technique [0002] The technology and products currently monitoring the potential risks of the elderly have the following features and defects: [0003] 1. Single and methods of product features are simple: such as resetting infrared detectors in the elderly toilet to detect whether the elderly is too long. [0004] 2. Defilter can monitor some of the physical quantities of the elderly, such as heart rate or fall. But in the actual use, the old man is reluctant to use wearable equipment 24 hours, which makes the penetration rate of wearable devices to be quite low so far. [0005] 3. Perceived the sensing technology and algorithms of the family's scenes are very complicated, immature, even if you use deep learning, you need expensive hardware settings. Therefore, the user's use cost is...

AI Technical Summary

Problems solved by technology

However, in terms of actual usage, the elderly are unwilling to use wearable devices 24 hours a day, which makes the penetration rate of wearable devices quite low so far

[0005] 3. The sensing technology and algorithm for perceiving the scene in the home are complex and immature, and even using deep learning requires expensive hardware settings

Therefore, the user's use cost is high

[0006] 4. The perception dimension is single, which cannot fully reflect and predict the possible risks of the elderly at home

For example, it can only make some simple judgments on the behavior of daily activities within a limited range, and cannot monitor areas beyond the range, and the hardware settings are also very complicated and expensive, and the practicability is poor

[0007] 5. All existing technical solutions are based on sensing what has happened to the elderly, and cannot predict potential risks

Images