How far can face mask detection systems help to curb COVID-19 spread

Accubits Technologies Inc
6 min readJul 23, 2020

--

Automated Face mask detection systems use AI models & deep learning algorithms to detect people who are not wearing face masks at public places.

Face mask detection using AI

Since COVID-19 was identified in December 2019, it has claimed thousands of lives and disrupted an even greater number of livelihoods. The spread of a pandemic can cause adverse effects on an economy through multiple channels, this involves both short-term fiscal shocks and long-term negative shocks that could disrupt economic growth.

Recent researches show that 40 to 60 million people are expected to be pushed into extreme poverty in 2020 — that is, living on less than $1.90 a day — as a result of the pandemic. Countries around the world have taken several steps such as extensive lockdowns, closing borders, enforcing social distancing, etc to reduce the spread of infection.

In June, the World Health Organization (WHO) suggested the widespread use of face masks as a way to prevent corona virus transmission. A study by a team of researchers led by a Texas A&M University professor demonstrated that not wearing a face mask drastically increases a person’s chances of being infected by the COVID-19 virus. From trends and mitigation procedures in China, Italy, and New York City, the researchers found that using a face mask reduced the number of infections by more than 78,000 in Italy from April 6-May 9 and by over 66,000 in New York City from April 17-May 9.

This has encouraged governments from all around the world to make the use of face masks mandatory in public areas. However, this move has been met with dissent from many. In fact, numerous protests against the wearing of face masks have broken out.

In such a scenario, how do government authorities and property owners enforce the safety standards required to curb the spread of COVID-19?

  • Through proper room ventilation, maintaining good hygiene, and proper use of disinfectants with an alcohol content above 70%.
  • Enforcing strict social distancing measures.
  • Enforcing lock-downs.
  • Effective quarantining of infected personnel.
  • Analyzing route maps of the infected personnel to predict the possibility of further infections in the vicinity.
  • Maintaining a well detailed medical record of patients at hospitals.
  • Restricting the travel path of peoples in high-risk zone areas.
  • By sending out frequent updates on COVID-19 pandemic transmissions.
  • Preventing mass gathering of people in public places.

But let me ask you,

Is it really possible to enforce these measures all the time?

Quite difficult right?

And that’s exactly why need a pandemic management system to tackle this situation.

Face mask detection systems using AI

Many companies are now working on face mask detection systems that use artificial intelligence to automatically detect whether a person is wearing a face mask or not. Most solutions have a similar concept that involves acquiring relevant data, training that AI model and finally integrating it into operations. However, there are still some problems associated with these solutions. Many of them require businesses to install additional hardware like a graphical processing unit, video management system, etc. Additionally, these solutions are not customizable and are unable to generate reports.

This is why we developed a face mask detection system that could overcome these hurdles and provide a more sophisticated solution for the accurate detection of face masks. This system can be easily integrated into pre-existing CCTV infrastructure for automated surveillance. Additionally, it will generate alerts to the concerned authorities whenever a person without a face mask is identified.

Some of the key features of the detection system include-

  • Easy to implement in any existing organizational system.
  • Customized alerts can be sent to the identified person without the face mask.
  • Can be used easily with any camera or hardware like surveillance cameras.
  • No need to install any hardware as the system can be connected with your existing surveillance system only.
  • Restricts access for those not wearing the masks and notifies the authorities.
  • Easily customizable based on your business requirements.
  • Allows for viewing of analytics based on the system generated reports.
  • Easy to access and control the movements from any device through face mask detection applications.

But how does the system work?

Like other solutions, it does not require any external hardware or specialized components to make the system work. It can be easily integrated with the existing CCTV infrastructure at your premise. It is primarily comprised of three main components that come along with it, they are as follows-

Data Interface

facial data interface
facial data interface

The system identifies people without personal protective equipment like face masks, face-covering equipment through real-time video analytics. The system’s interface can store information on previously identified faces, phone numbers, and e-mail addresses. This information can be used to send people alerts if they forget to wear a mask. On identification of a new face, alerts can be sent out to the concerned authorities.

Surveillance Dashboard

surveillance dashboard
surveillance dashboard

The system provides a dashboard that shows the statistics of face mask violations identified, alerts generated, etc. The system automatically notifies the concerned security officer to take defined actions upon detection of violations.

Notification Interface

notification interface
notification interface

A mobile application can be used to receive alerts whenever the system detects anyone who isn’t wearing a face mask. Notifications can be programmed to pop up on the surveillance dashboard in real-time as well.

Also, the system can provide you up to 95–98% accuracy on the detection of face masks which can be further enhanced depending upon the capability of the CCTV camera used at your premise.

Applications of the face mask detection

In the current state of the world, such a system can be beneficial in a number of scenarios. It can be installed in various public areas to enforce safety standards and social distancing requirements as well. A few examples of where we could use face mask detection are-

  • Airports: Cross-border travel is one of the leading reasons for increased infection. This is why it is important to hook up a face mask detection solution in the frequently used parts of the airport like immigration areas, boarding gates, etc. The system can alert the authorities whenever it detects someone without a face mask. It can also send notifications directly to the airport staff if they are seen without the required personal protective equipment.
  • Offices: Employers must ensure that their workers are following the required safety standards. They can use the systems to ensure that their employees are taking precautionary measures to stay safe. Reports can be generated to identify staff members who are not following the safety requirements.
  • Hospitals: Hospitals can be considered as hubs for infection due to the sheer number of sick patients present. Although most hospital staff usually wear masks at all times, the same cannot be said about patients. Hospital authorities can be alerted if a patient is seen without a mask so that corrective action can be taken immediately.

By helping government authorities and property owners ensure that everyone is taking the required precautionary measures, a reliable face mask detection solution can reduce the spread of the pandemic to a large extent. Additionally, direct contact can be established to the public through notifications that remind them to wear their face masks at all times.

Let's see how to face mask detection works on ARM-based devices:

Face mask detection on arm based devices & other embedded systems

Chances for improvement

As observed from the real-time face mask detection video, the solution is the perfect choice for enforcing strict social distancing measures at government organizations, offices, health organizations, public places, etc.

To improve the level of accuracy of the detection system even further,

Firstly it should be trained in different scenarios at your premise. As there’s no replacement for the real-time data.

Secondly, during the data gathering process, the unique object classifier of the system should be put in situations where the face of the detected person is covered with clothing rather than a face mask, etc.

To get better insights into how to face mask detection system helps to keep your workplace safe, this article would be useful:

--

--

Accubits Technologies Inc

Accubits Technologies is an enterprise solutions development company focusing on AI and Blockchain technologies, based in Virginia, USA. https://accubits.com/