COVID-19 Early Detection System
College of Engineering
Author
05/03/2021
Added
30
Plays
Description
We have designed a passive device that can identify and take the temperature of people's foreheads while they walk through a doorway. Given the current pandemic that we have been facing, we wanted to create a low-cost solution to prevent the further spread of COVID-19. The goal of our project is to collect enough data from our device to help characterize buildings or any place with large gatherings of people. This would allow epidemiologists to quickly identify local outbreaks or high-risk areas and assist in contact tracing. At the heart of our project is a microprocessor running Linux. Combining our microprocessor with an image sensor and medical-grade infrared thermometer, equipped with a five-degree field of view lens, allows us to detect and take temperature readings of a person's forehead.
Searchable Transcript
Toggle between list and paragraph view.
- [00:00:05.511]The COVID-19 Early Detection System
- [00:00:07.870]is a public health tool that aids epidemiologists
- [00:00:10.840]and researchers in measuring
- [00:00:12.070]COVID-19 infection rates across populations.
- [00:00:16.980]Currently, many infection rate models use data
- [00:00:19.610]from individual COVID-19 tests
- [00:00:21.710]to determine general population infection rates.
- [00:00:25.330]While these individual tests are extremely accurate,
- [00:00:28.480]it is not easy to collect this data into a model reflective
- [00:00:32.120]of the actual infection rates of a population.
- [00:00:35.340]Few people actually get tested, and if they do get tested,
- [00:00:38.720]they don't represent a statistically random sample.
- [00:00:42.780]The COVID-19 early detection system is a low-cost,
- [00:00:46.110]forehead temperature measurement device
- [00:00:48.370]that can be deployed en masse
- [00:00:50.070]in high traffic, public places.
- [00:00:55.470]The naive solution to this project
- [00:00:58.000]would be to use a thermal camera
- [00:00:59.570]to identify all the foreheads in a frame,
- [00:01:02.220]and then catalog that data.
- [00:01:04.370]However, these thermal cameras are expensive.
- [00:01:08.450]Instead, the COVID-19 early detection system
- [00:01:11.480]uses a low-cost, single element infrared thermometer,
- [00:01:15.030]and waits for a forehead to move into the field of view.
- [00:01:18.290]When deployed on a massive scale in high traffic areas,
- [00:01:21.790]it provides enough data to make good statistical inferences
- [00:01:25.390]about infection rates.
- [00:01:28.920]This system uses an extremely low-cost,
- [00:01:31.500]consumer-grade image sensor,
- [00:01:33.400]and a sophisticated face detection algorithm
- [00:01:36.060]to determine when a forehead
- [00:01:37.480]is in the IR temperature sensor's field of view.
- [00:01:40.860]The system runs embedded Linux,
- [00:01:42.610]which provides out-of-the-box support
- [00:01:44.100]for open CV image processing, 802.11 networking,
- [00:01:49.430]and multimedia streaming.
- [00:01:51.630]This system is built around the NXP i.MX6,
- [00:01:55.200]which hosts a single core Arm Cortex-A7
- [00:01:58.780]and features all the necessary peripherals
- [00:02:00.840]required for the project.
- [00:02:02.950]The Allwinner A33 was also considered,
- [00:02:05.740]though it is more powerful than necessary.
- [00:02:10.380]After proving out the image sensor
- [00:02:12.340]and IR temperature sensor on an i.MX6 breakout board,
- [00:02:16.010]a first revision PCB was designed.
- [00:02:18.590]This first spin of boards was meant to better interface
- [00:02:21.620]with the sensors and parts.
- [00:02:25.330]The first revision hardware
- [00:02:26.770]included a five-inch TFT,
- [00:02:29.050]but this added unnecessary costs
- [00:02:31.010]and complexity to the design
- [00:02:32.930]while not offering additional features or capabilities.
- [00:02:37.330]The second revision of the board
- [00:02:39.050]added a wifi and Bluetooth module,
- [00:02:41.520]and changed the layout to better fit
- [00:02:43.520]into a 3D printed enclosure
- [00:02:47.840]For the color camera,
- [00:02:49.110]the product uses the OmniVision OV7670 Image Sensor.
- [00:02:54.020]While this is an older 640 by 480 resolution design,
- [00:02:57.860]this provides sufficient resolution for facial tracking.
- [00:03:01.230]And these image sensors are dirt cheap,
- [00:03:03.670]as low as 30 cents each, which helps drive down the cost
- [00:03:06.730]of the product, important for massive scale installations.
- [00:03:11.590]This was prototyped with an off-the-shelf camera module
- [00:03:14.570]to test the Linux kernel drivers support
- [00:03:16.950]before the actual image sensor was integrated
- [00:03:19.670]into the PCB design.
- [00:03:23.960]The product uses the Melexis MLX90614 DCI
- [00:03:28.550]medical grade, IR temperature sensor,
- [00:03:31.170]equipped with a five-degree
- [00:03:32.410]field-of-view lens
- [00:03:33.370]to get more precise readings at long ranges.
- [00:03:38.070]Once the system identifies a person's forehead
- [00:03:40.740]in the field-of-view of the IR thermometer,
- [00:03:43.610]a reading is collected, and sent to a web server
- [00:03:46.400]using a standard HTTP rest API.
- [00:03:50.980]This web server software
- [00:03:52.410]was written in ASP.NET Core with Entity Framework
- [00:03:56.100]and Razor Web Views.
- [00:03:59.090]The goal of the industrial design of the product
- [00:04:01.430]was to have a fun, inviting appearance.
- [00:04:04.230]Several 3D prints helped explore the ideas
- [00:04:06.880]of its form and function.
- [00:04:09.130]Painted in bright purple,
- [00:04:10.650]the product has a circular, bubbly image
- [00:04:12.960]that looks friendly and inviting.
- [00:04:15.370]On the back, a standard 1/4-20 threaded insert
- [00:04:18.500]enables mounting using off-the-shelf camera hardware.
The screen size you are trying to search captions on is too small!
You can always jump over to MediaHub and check it out there.
Log in to post comments
Embed
Copy the following code into your page
HTML
<div style="padding-top: 56.25%; overflow: hidden; position:relative; -webkit-box-flex: 1; flex-grow: 1;"> <iframe style="bottom: 0; left: 0; position: absolute; right: 0; top: 0; border: 0; height: 100%; width: 100%;" src="https://mediahub.unl.edu/media/16790?format=iframe&autoplay=0" title="Video Player: COVID-19 Early Detection System" allowfullscreen ></iframe> </div>
Comments
0 Comments