|
Face and Hand Gesture Recognition System Using MediaPipe |
|---|---|
| รหัสดีโอไอ | |
| Creator | Sethakarn Prongnuch |
| Title | Face and Hand Gesture Recognition System Using MediaPipe |
| Contributor | Tanawat Sangkhakul, Nutchanon Angchoun, Wuttipong Pinsarmsak, Kant Charoenjit |
| Publisher | Faculty of Engineering and Industrial Technology, Kalasin University |
| Publication Year | 2568 |
| Journal Title | วารสารวิศวกรรมและเทคโนโลยีอุตสาหกรรม มหาวิทยาลัยกาฬสินธุ์ |
| Journal Vol. | 3 |
| Journal No. | 4 |
| Page no. | 45-58 |
| Keyword | Face Recognition, Hand Gesture Recognition, OpenCV, MediaPipe, SciPy |
| URL Website | https://ph03.tci-thaijo.org/index.php/JEIT |
| Website title | วารสารวิศวกรรมและเทคโนโลยีอุตสาหกรรม มหาวิทยาลัยกาฬสินธุ์ |
| ISSN | ISSN 2985-0274 (Print),ISSN 2985-0282 (Online) |
| Abstract | Current recognition systems face challenges related to accuracy, real-time speed, and limited device resources. This paper presents a face and hand gesture recognition system using MediaPipe. The research aims to create a real-time face and hand gesture recognition system utilizing MediaPipe, OpenCV, and SciPy, running on a Raspberry Pi 4 board with a webcam. The experiment involved a sample group of 100 participants. The experimental results show that the system achieves a success rate of approximately 95% in unlocking a security door for individuals whose information is stored in the database, despite limitations in camera angles and lighting. The highlight of this research is its presentation of a dual authentication method using the cosine similarity technique for security systems. The system’s performance and accuracy will be further evaluated in real-world scenarios, such as its application in educational institutions within the Thai context. |
