The Ability Test of Model the Facial Landmark Detecting

Main Article Content

Petchara Inthanon
Surasak Mungsing


Recently, there are several techniques for face detection such as detecting the whole face or detecting specific parts of face by using artificial intelligence. For this project, facial landmarks model was conducted to detect 68 points of different face parts. Each facial point could be calculated to generate values. Moreover, various face motion photos were applied into 1,522 datasets to examine the accuracy of the facial landmarks model. The result showed that the accuracy of using model was very high which was 94.81%. Besides, the model was conducted to detect unusual long closed-eyes and opened-mouth by comparing 2 tools between Nvidia Jetson Nano and Raspberry Pi 4 + Intel® Neural Compute Stick 2 with webcam camera. There were 4 circumstances to test the detection of facial points and the average accuracy of each tool was 78.08% and 80.59% respectively.

Article Details

How to Cite
Inthanon, P., & Mungsing, S. (2020). The Ability Test of Model the Facial Landmark Detecting. PKRU SciTech Journal, 4(2), 13–22. Retrieved from
Research Articles


Soukupová, T., & Cech, J. (2016). Eye blink detection using facial landmarks. (Research Reports). Slovenia: Czech Technical University in Prague.

Taufique, A. M. N., Savakis, A., & Leckenby, J. (2019). Automatic Quantification of Facial Asymmetry Using Facial Landmarks (pp. 1-5). In 2019 IEEE Western New York Image and Signal Processing Workshop (WNYISPW). Rochester, New York.

Wei, W., Tian, C., & Zhang, Y. (2015). A two-stage facial landmark localization method (pp. 157-160). In 2015 International Conference on Orange Technologies (ICOT). Hong Kong, China.

พรพิมล ตินะน้อย และ ณิชมน ทองพัฒน์. (2562). มุ่งเป้าปลอดภัย ‘จยย.’ ลดตายติดอันดับโลก. [ออนไลน์], สืบค้นจาก (30 มีนาคม 2563).

กลุ่มสถิติการขนส่ง กองแผนงาน กรมการขนส่งทางบก. (2562). รายงานสรุปผลอุบัติเหตุรถโดยสารสาธารณะ ประจำปีงบประมาณ พ.ศ. 2562. [ออนไลน์], สืบค้นจาก statistics/ (25 มีนาคม 2563).

Wong, J. Y., & Lau, P. Y. (2019). Real-Time Driver Alert System Using Raspberry Pi. ECTI Transactions on Electrical Engineering, Electronics, and Communications, 17(2), 193-203.

Terada, T., Chen, Y. W., & Kimura, R. (2018). 3D facial landmark detection using deep convolutional neural networks (pp. 390-393). In 14th International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery (ICNC-FSKD). Huangshan, China.

Huber, P. (2017). Real-time 3D morphable shape model fitting to monocular in-the-wild videos (PhD Thesis). UK: University of Surrey.

Khunthi, S., Saichua, P., & Surinta, O. (2019). Effective Face Verification Systems Based on the Histogram of Oriented Gradients and Deep Learning Techniques (pp. 1-6). In 14th International Joint Symposium on Artificial Intelligence and Natural Language Processing (iSAI-NLP). Chiang Mai, Thailand.