Design and Build a Prayer Rak’ah Reminder Device for Elderly People with Pose Detection Using MediaPipe Based on Raspberry Pi

Abstract

Establishing the five obligatory prayers is a necessity that Muslims must undertake. Problems often occur in people with memory problems, such as the elderly. Obstacles that often occur include forgetting the rak'ah and difficulty remembering the next pose to be performed. New technologies continue to emerge including digital imagery. Digital imagery can be used to help with these problems by utilizing pose detection using the MediaPipe library. MediaPipe is used to determine body parts visibility and joint angles captured by the webcam to detect performed pose. By detecting the pose, the output is then generated in an LED Matrix display namely the rak'ah and pose. The results of this study showed that the percentage of success in identifying ruku’ is 93.73%, i’tidal is 94.12%, sujud is 92.55%, first tahiyah is 89.17%, final tahiyah is 82%, with the highest percentage of 98.04% in standing pose. The pose detection success percentages based on the distance between the performer and the webcam are from 150cm is 91.88% success percentage, at 200cm success percentage is 92.42%, and at a distance of 250cm is 93.75%, with the highest success percentage at the distance of 250cm. The system average delay for detecting poses is 1.028 seconds.

Keywords

• prayer rak’ah
• prayer poses
• mediapipe
• pose detection