Detecting fever on the crowd

Our project helps to detect the presence of sick individuals in a locality (schools, malls, churches, public squares, parks, etc.)  through the use of an electronic surveillance thermal imaging system that captures images of the crowd and detects individuals with high body temperature (fever). In addition to accounting data for statistical analysis, the system also allows the individual to be contacted for health assistance.

The surveillance system keeps the individual privacy and only publish online anonymous data for  statistical analysis. Therefore,  it helps the community to detect and early respond  to emergence of diseases.

Nowadays, all nations struggle to deal with a pandemic disease by isolating people, testing the population while seeking for a cure. However, the short history shows us that it was not first nor will be the last pandemic during our life-time. Communities, states and countries need to monitor and detect as early as possible the emergence of diseases on the local population to be able to respond the health issue.

Many times a higher level health agency cannot detect an outbreak  until it is too late. Therefore, we propose  an embedded surveillance system to be installed in many points in a city. The surveillance system is always working detecting people on the crowd and analyzing signs of individuals with fever and keep data on an online database. With a network of surveillance systems updating a health database on the cloud we decrease the reaction time to detect and react to a threat. 

The solution is based on an embedded electronic system (small computer) with a thermal camera. The embedded device captures frames in real-time, detect individuals on scene using image detection algorithms and also their body temperatures. As the device is connected to the Internet it updates a database on the cloud with every detection and also dispatches an inquiry alert by SMS message or push notifications to the nearby authorities.

The solution helps to decrease the detection and also the reaction time to an emerging disease among the population. Specifically in situations like this it has an important impact.

Our solution is a tool to detect an outbreak of a disease by capturing the images of a population. It allows the authorities to start working against the heath threat in advance and then  save financial resources and probably many lives. 

It has becomes a normal scene to see police officers and health assistance using hand thermometers trying to detect people with fever on the crowd. This method is inefficient, expensive and also hurts the individual privacy as many  people can feel humiliated by been exposed publically to this procedure. Even worst in situations when the detection is positive.  Our solution is an "invisible" (in the sense it does not disturb the crowd)  device which automatically detects people on the crowd and alerts the heath assistant's to approach the individual for a conversation and deeper health check. Our proposal provides a more human approach to handle a threat in a World suffering from traumas and fear.

application of an existing technology. 

1) Raspberry Pi 4, thermal camera

 2) embedded software (tensor flow) with customizations for this specific problem.

There are plenty of examples of use of this technology to recognize faces or objects. Our solution uses the same technology with a different goal in mind.

1) https://www.youtube.com/watch?v=Rgpfk6eYxJA

2) https://www.youtube.com/watch?v=aimSGOAUI8Y

3) https://www.youtube.com/watch?v=r0h2aNqXbBw

4) https://www.youtube.com/watch?v=2ddii2Gnti8

5) https://www.youtube.com/watch?v=MsylEq6tTeE

I believe our solution has an impact as it  provides a continuous surveillance  24/7 and helps to early detect  an outbreak of a disease. Also, it minimizes the impact on the population which do not need to be subject to the invasive and inconvenient temperature checks by police of health assistants (that leads to a  panic state and should be avoided)