Acoustic Gaze: How Sound Waves Allow "Seeing" Through Walls - Alternative View

Acoustic Gaze: How Sound Waves Allow "Seeing" Through Walls - Alternative View
Acoustic Gaze: How Sound Waves Allow "Seeing" Through Walls - Alternative View

Video: Acoustic Gaze: How Sound Waves Allow "Seeing" Through Walls - Alternative View

Video: Acoustic Gaze: How Sound Waves Allow
Video: If You See Square Waves In The Ocean Get Out Of The Water Immediately 2024, July
Anonim

Chinese experts have created an ultrasonic device for monitoring human movements. The device is able to "see" through walls, recognize different body positions, and follow the movements of the object. The developers believe that such an acoustic recognition system has a number of advantages over video cameras and electromagnetic radar detectors.

Chinese scientists from Wuhan University of Technology have created a device that is capable of detecting a person using ultrasound and fixing the position of his body, even if he is behind an obstacle out of sight. This is reported in an article published in the journal Applied Physics Letters.

For military and civilian purposes, radar detectors are often used to detect and monitor people, which, using electromagnetic radiation, are able to detect objects out of sight, but such devices are not able to recognize the position in which a person is.

To create their own observation system, specialists from China decided to use an alternative environment - acoustic. Their device "sees" what a person is doing using ultrasound.

“The possibilities of using the device will expand as the accuracy of the recognition system increases. For example, it can function as a home medical alarm. The system is capable of detecting that a person has fallen at home, and assistance will be promptly provided to him,”said the developer of the acoustic device, associate professor at Wuhan University of Technology Xinhua Guo.

The device in the form of an acoustic lattice is equipped with 256 receivers and four ultrasonic transmitters, which are included in a convolutional neural network - a self-tuning system for signal analysis and object detection. The speaker system operates at an ultrasonic frequency of 40 kHz to ignore potential noise interference. Currently, she is able to distinguish between four static and dynamic states of the human body: standing and sitting positions, as well as walking and falling.

“In the future, we will improve the system for recognizing the positions of the human body, including arbitrary postures, in order to eliminate possible errors. As you know, the motor processes of the human body are quite complex. People can move in different ways, assume different positions and fall. We plan to study better the mechanics of human fall in order to achieve greater accuracy of the device,”says Xinhua Guo.

According to him, the developers are currently investigating the possibility of using the system in various devices for commercial and home use.

Promotional video:

Author: Arseny Skrynnikov

Recommended: