System and method for detecting abnormal status of user based on bio-signal and wearable device

High-accuracy biosignal wearable device

This technology relates to a system and method for detecting user abnormalities based on biometric signals and wearable devices, which can prevent accidents in advance by classifying and detecting abnormal conditions of the driver in advance based on biological signals measured using the device and notifying them to the driver and passengers.

The biosignal-based method has the inconvenience of having to attach a measuring device to the driver's body to collect biosignals, so a wearable device was designed, but the existing method is based on the measurement device. Because it is very sensitive to movement and contains a lot of noise, there is a high possibility of incorrect status judgment when used for status determination due to the inaccuracy of the acquired signal. To solve this problem, this technology proposes a method of using Adaboost, one of the ensemble learning techniques that produces the final result.

This technology is capable of detecting abnormal conditions of the driver in advance and notifying them to the driver and passengers. In addition to preventing traffic accidents, the condition detection algorithm can be applied to various products such as various healthcare products or systems that monitor the condition of workers in other industrial sites.

‍

‍

‍

Key Features:

• After measuring several signals from the driver, inspecting the measured signals and performing a pre-processing process to check whether the signals can be used to detect abnormal conditions
• In order to have high-accuracy detection performance, use Adaboost, an ensemble learning technique that combines the results of multiple classifiers to produce the final result
• The driver's biological signals such as PPG signals, temperature, and GSR signals, and acceleration and gyro signals Measuring signals related to movement
• Designing a neural network-based classifier and combining the results of several trained classifiers to perform final state classification and detection

‍

‍

This technology was developed through support from the National IT Industry Promotion Agency's bio-signal-based driver abnormality detection and notification system research project.

‍