Authentication System by using HOG Face Recognition Technique and Web-Based for Medical Dispenser Machine
Lukman Chaiyarab, Chaiyanas Mopung, Theekapun Charoenpong
Abstract
An infectious and contagious disease affects human life significantly. Unexpectedly, monthly check-up of patients is inconvenient to receive a medical supply. Conveniently receiving medical supplies is a solution to hospital patients, eliminating crowded environments, and reducing contamination of the surface area. Therefore, this paper proposed a medical dispenser machine composed of two parts which are Web-based and face recognition in real-time. The medical profession can adjust the quantity of medicine. After the pharmacist arranged medicine, the email is distributed to the patients with the date and time for receiving packages. An automated system with a machine-learning algorithm assists them in receiving medical supply by using face recognition with Histogram Orientated Gradients (HOG) embedded in Raspberry Pi 4B. Face recognition is the ultimate technology to improve and make an impact in people's lives by diminishing contamination by viruses by not touching any surface area. The result of 50 image inputs showed an impressive recognition of authorized individuals with an accuracy of 80.0 %. This work combines hardware set up with PHP Web-based to distribute the medical supply efficiently. Subsequently, the implementation performed a satisfying performance that was suitable for the hospital.