Embedded Files
Wearable Fever Screening System
Wearable Fever Screening System
Spring 2020 MAE 156B Sponsored Project
University of California, San Diego
Sponsored by the UCSD Electrical & Computer Engineering Department
Background:
COVID -19 is a pandemic that is seriously impacting the economy and civilian life. It is essential to maintain public health by detecting who is ill and preventing the spread of the virus. For this reason, Dr. Wang and his team from UCSD Electrical & Computer Engineering department are working closely on developing a pair of thermal imaging glasses which is able to detect the thermal signature of the human body at a certain distance and send out a signal if the infrared sensor detected a person who has a fever. The invention of this device will make a great social impact by easily detecting the people who are sick. It will allow screeners to detect high temperatures at a distance and potentially screen large quantities of people quickly. With the help of the device, it will assist in identifying high-risk individuals so that they can be referred to medical facilities in order to reduce the rate of infection.
Summary of Project Objectives:
There are several objectives in this project:
To test the capabilities of cameras in different environmental conditions
Accuracy of sensing the body temperature
Range of distance from the target
Ability to distinguish background and the human body
Range of temperature it detects
2. Calibrate the thermal sensor into the most optimal working condition
3. Design a wearable system for both the thermal sensor camera and the power source and connect it with the glasses.
Final Design:
The final design of this product is shown below. The MLX90640 with 55° FOV is the chosen camera for its high accuracy and precision at a distance range of approximately 2 meters. The camera case is paired with different attachments using a dovetail design. These attachments can be used to mount the camera to different sized glasses or a hat. The camera is connected to a Raspberry Pi located at the waist, along with the power source, which is a 26800 mAh phone power bank with two USB outlets. Raw data from the camera is processed, and a normalized image is displayed to smart glasses, which will also be connected to the Raspberry Pi, either through a USB cable or Bluetooth depending on the battery life of the smart glasses.
The camera is programmed at a refresh rate of 2 Hz with built-in noise cancellation. The user must input ambient temperature to the device for maximum accuracy, as ambient temperature has a large effect of measured temperature. Additionally, accuracy is optimal when the target is located in the target zone specified by the red or green box on the screen at a distance of 1.5 to 2.5 meters away from the user. Other device specifications are listed in the table below.
Summary of Performance Results:
Link to Executive Summary: https://docs.google.com/document/d/1DsKCoSu4lDe1Jpi4-rKLvcszcpJkWWKJpi1odcdiqMM/edit
References
Figure 1: https://www.dmagazine.com/healthcare-business/2020/03/coronavirus-has-arrived-in-dallas/
Figure 4: Reference: https://www.scmp.com/tech/gear/article/3077122/hangzhou-park-security-uses-ai-powered-smart-glasses-detect-people-fever
Page updated
Report abuse