Fighting Pandemic With Computer Vision

The global Covid-19 pandemic is officially one year old. As reported on the covid.19.go.id website, Indonesia has reached more than a million positive cases per February 4, 2021. Meanwhile, globally, there are more than one hundred million cases per February 4, 2021. It is safe to say that this pandemic is still a long battlefield to face.

Therefore, all kinds of capacities are needed to fight the spread of this virus. One of the capacities is the use of technology. In this article, we will specifically discuss how computer vision, as a branch of AI, can help the process of dealing with a pandemic. Quoting from Ulhaq et. al (2020), computer vision is recently favored for solving various complex problems in health care and has the potential to contribute to the fight of controlling COVID-19.

For example, let’s take a look at the Taiwan Airport case. According to the IEEE Computer Society, regardless of Taiwan’s proximity to the epicenter of the coronavirus outbreak, Taiwan has reported fewer cases of the coronavirus. The center of the marvelous crisis management is the country’s ability to include the use of temperature monitoring in airports to screen travelers for fever through computer vision.

Other than thermal detection, computer vision can be also used to detect social distancing and mask using. As written on Brouton Lab, a deep learning technique enables the calculation of the distance between people. For mask use, computer vision has two tasks in mask detection such as object detection with neural networks and classification of faces which is also with a neural network.

Business-wise, computer vision can also be utilized for people counting. The occupancy rate in a building might be a hard thing to identify. Business owners might have put their employees at risk, considering that the recovery of this Covid-19 pandemic might take a long time. 

In order to get the best out of the feature, business owners should determine the ideal number beforehand. This standard can be adjusted in accordance with the local or national public policy. Using this standard, the system then can give a warning if there is an occupancy number violation.

However, computer vision implementation faces some challenges. Brouton Lab writes that there are three challenges that might occur. First is the potential camera to be installed that still serves in a 2D way. Secondly, person identification issues might happen if the machine can not identify the right person. Lastly, if there are a few cameras being connected in one network, following someone across those cameras might be hard to be done.

There are no one-size-fits-all types of solutions. Yet, unprecedented times call for radically smart solutions. So, feeling the need to answer the call, Qlue presents a solution called QlueVision. QlueVision is a computer vision technology equipped with an IP-camera like CCTV, webcam, and smartphone. 

QlueVision has generated more than 5,5 million data of CCTV reports, from facial recognition, vehicle detection, object detection, and any other data. Utilizing QlueVision will enable business owners and/or government officials to allocate their on-ground personnels effectively. Moreover, equipped with an integrated dashboard, a data-driven policy to fight Covid-19 can be established quickly. 

With more than 98.5% accuracy, QlueVision provides notable savings with flexibility and scalability through automatic updates to save time by an easy to use cross-platform. In other words, smart viruses can be fought by implementing smart solutions and QlueVision is one of those smart solutions.

References:

• Ulhaq, A., Khan, A., Gomes, D., Paul, M. (2020). Computer Vision For Covid-19 Control: A Survey.
• https://www.computer.org/publications/tech-news/covid19-research/thermal-detection
• https://broutonlab.com/blog/computer-vision-to-fight-pandemic