AI non-sensing temperature solution

2020-12-22 15:21:42 fandoukeji

As various places usher in a large-scale return to work, the prevention and control of the new coronavirus pneumonia epidemic has also entered a critical period. In order to win the fight against the epidemic, key areas such as airports, railway stations, subway stations, schools, and large communities have initiated body temperature testing to strengthen the screening of persons with abnormal body temperature and suspected patients. Relying on the accumulation of leading artificial intelligence technology, the 6108 solution launched the ‘AI epidemic prevention’ solution for the first time. The product adopts a non-contact method to improve the efficiency of large passenger flow places and reduce the risk of infection of staff. At the same time, by combining the leading AI algorithm with thermal imaging temperature measurement technology, the accuracy of body temperature detection reaches ±0.3, which meets the criteria for preliminary screening of suspected patients. In addition to the temperature measurement function, the product also supports identifying and prompting whether a person is wearing a mask.




"Infrared temperature measurement products" related artificial intelligence technology solutions, hoping to achieve "quickly realize the purpose of rapid detection of large-scale high-temperature groups." For this reason, the 6108 program proposes "personnel temperature + mask wearing + personnel identity" and other aspects of identification and management AI epidemic prevention solutions.




AI non-sensing temperature solution






   It is reported that the plan consists of a regional access module and an entrance and exit access module, which can uniformly display and archive real-time information. Among them, the regional traffic module can realize non-contact screening in multiple scenarios such as airports, stations, subways, schools, office buildings, communities, etc., to achieve high temperature, abnormal events without masks and personnel identification.




   After deploying this module, the system uses a thermal imaging camera to measure the forehead temperature of pedestrians. It can automatically detect people whose body temperature exceeds 37.3°C in the crowd, and generate sound and light/pop-up reminders. The staff can further use the medical thermometer to make the second measurement and then check and confirm in the intelligent management system.




   In addition, the system also supports identification reminders for persons not wearing masks, etc., and can identify persons with high accuracy while wearing masks. Through the above methods, management personnel are provided with a full range of epidemic prevention information assistance such as the body temperature of the personnel, the wearing status of the mask, and the identity of the personnel.




AI non-sensing temperature solution






  The entrance and exit pass module of the program combines face recognition, non-sensing temperature measurement and non-sensing passage, which can quickly identify personnel information and perform temperature measurement registration, while providing background records.




   The entrance and exit pass module of the new system is used to accurately identify employees without removing their masks and record their attendance. In the case of personnel wearing masks, the pass rate can reach 85% when 50% of the bridge of the nose is exposed. At the same time, masks can also be set to identify, and employees must wear masks to enter the office area.




   The regional access module of the AI smart epidemic prevention solution consists of an intelligent temperature measurement and screening terminal, a thermal imaging camera, a black body and an intelligent temperature measurement and screening system. It can optimize the accuracy of thermal imaging temperature measurement through artificial intelligence algorithms, and achieve high-precision temperature detection and screening. At the same time, the solution team is still optimizing the algorithm. In the future, it will be possible to achieve ±0.3 accuracy even further without breaking the black body.