Tongue Biometrics Extraction Based on YOLO Algorithm and CNN Inception

Meo Vincent C. Caya, Miguel Ephraim C. Caringal, Khristopher Allen C. Manuel

20212021 IEEE 13th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)26 citationsDOI

Abstract

Biometrics is the measurement of the uniqueness of the person’s physical characteristics. The tongue has its unique features. The texture, shape, and color of the tongue are different from one another. This paper envisions the development of Tongue Print Biometrics. This study is accomplished by providing the system with data captured by the Camera Module that is connected to the Raspberry Pi. The images captured are trained using two algorithms: the YOLOv2 and Inceptionv3. The process is divided into two parts: Enrollment and Testing while the Training part was included in the enrollment phase. Based on the results, the first algorithm which is YOLOv2 yields a rate of 98.8% accuracy while the second algorithm which is Inceptionv3 yield a rate of 39.44% accuracy.

Topics & Concepts

BiometricsComputer scienceArtificial intelligenceTongueComputer visionFeature extractionProcess (computing)Texture (cosmology)AlgorithmPattern recognition (psychology)Image (mathematics)Operating systemLinguisticsPhilosophyTraditional Chinese Medicine StudiesInfant Health and DevelopmentHand Gesture Recognition Systems