<![CDATA[COVID-19 is a disease caused by the corona virus and it has caused a global pandemic since December 2019. The use of masks is highly recommended to stop the spread of the corona virus, especially in closed rooms with many people inside, such as the classrooms and the office spaces. Detection of people with mask to access door based on the ratio of face Bounding Box and Roundness using Naive Bayes on Raspberry Pi is used to ensure people wearing masks who want to enter classrooms and office spaces. Webcam is used to capture images of people who want to enter. The image is processed and classified on the Raspberry Pi 4. Image processing begins with converting RGB to YCbCr and performing the morphological dilation, the morphological opening, and the morphological closing of morphology. Image processing aims to segment the human faces and remove their backgrounds. Human facial features were extracted using the ratio of Bounding Box and Roundness analysis which aims to determine the detected human face. The method for classifying faces is the Naive Bayes method. The solenoid lock opens when the classification result uses a mask, and it will be locked when the classification result does not use a mask. In the process of testing the Naive Bayes model using 60 data, the highest accuracy is 90%. To prove the accuracy of the Naive Bayes model, a test was carried out by inputting images directly into the system at 5 different distances, namely at 0.5 meters, 1 meter, 1.5 meters, 2 meters, and 2.5 meters. The test at each distance got an average accuracy of 86.6%. The average execution time required for system to detect masker is 8.82412 milliseconds.]]>
Copyrights © 2021
