The well known epidemiological characteristics of COVID-19 still stress the importance of the urgent priorities to be taken into account.
In the first phase of Covid-19 pandemic, the health emergency worsened due to the very high infectivity of the virus which produced thousands of infections in a very short time. Italy just entered in Phase 2 of the coronavirus emergency, a transitory period of coexistence with the virus in which the Italian Government activated measures for a gradual return to the everyday life, with the reopening of some activities and partially allowing mobility in the cities.
If, on the one hand, the need for a return to normality is predominant both from a social and merely economic point of view, on the other, it will be crucial the activity of control and prevention of crowd gatherings so to allow the "economic awakening" of the social fabric while preserving the safety of the population.
From an epidemiological perspective, social distancing measures reduce the value of the effective reproduction number rate of the epidemic (R0): avoiding crowd gatherings and persevering with social distancing on a large-scale and medium-long term, seems the main road to follow to reduce the number of spreading events transmissions.
In addition to this fundamental assumption, while waiting for a pharmacological treatment of the virus, the scientific community identified the weapons to contrast and limit the spread of the virus by timely screening each subject and with the adoption of individual safety devices.
Reco 3.26 translated each of the virus countermeasures indicated by the scientific community into a technological dimension.
Reco Crowd Detector, Reco People Counting and Reco DPI Check are intelligent video surveillance software for the identification of crowd gatherings and situations of potential Covid-19 contagion and for workplaces safety. They are based on the most innovative Artificial Intelligence algorithms applied to Computer Vision, which exploit continuous and deep learning models through efficient and reliable Deep Learning and Convolutional Neural Network systems optimized for each use case.
These technologies can be used to prevent situations of potential COVID-19 contagion through a territorial control action, using video flow analysis systems coming from urban video surveillance cameras already existing in our cities and available to the Police Force or cameras set up in shops, public transport or mobile cameras. The crowd detection system is able to detect and report in real time the formation of crowd gatherings while people counting node allows to make a precise and real-time count of the number of subjects accessing or passing through the monitored video areas.
Reco DPI Check solution allows protecting the health and safety of citizens from possible coronavirus infection by verifying that the prescribed personal protective equipment is worn (medical masks, protective goggles, cap and high visibility protective clothing). Reco DPI Check can be integrated on tablets and mobile cameras to monitor and guarantee access to workplaces.
This real-time monitoring will trigger alerts to the command and control system, both when it detects an increase in the density of aggregation of people and when the meter exceeds the "safety threshold" set up by the operator (maximum number of eligible subjects). All the parameters mentioned above: density, monitored area, number of people allowed can be parameterized camera by camera in order to monitor an average distance between people.
The system will also allow to send periodic and customizable images to the operator in order to monitor the evolution of the situation or by reporting the percentage increase in the number of people over time, compared to the previous alert, so that the system generates further alarms or critical alarms. The alerting system can also trigger external systems: traffic light that turns red or green depending on the assembly in a specific room; automatic door that opens or not etc.
In addition to the previous solutions, Reco Fever Check uses thermal cameras to measure, with a minimum approximation, the body temperature of each subject present in the framed scene. The high level of accuracy is guaranteed by the use of the simultaneous combination of two Deep Learning algorithms operating on the dual sensor (optical and thermal) which the camera is equipped with.
The combination with Reco Face Detection component increases the screening performance, significantly reducing false alarms due to external heat sources, even for low-end thermal cameras.
In conclusion, we believe that the use of Reco solutions, combined with the sense of responsibility of each individual, will help to confine the transmission of the virus, flattening the epidemiological curve and allowing our National Health System to support the shock wave of daily infections and in this way, we can give it precious time to treat the most serious cases and, in the long term, to develop vaccines and pharmacological treatments.