Driver State Alert Control using Head Shoulder Inclination and Facial Landmarks
Ashitha Ebrahim1, Fathimathul Harshima P. T.2, Aby Abahai T.3
1Ashitha Ebrahim*, Department of Computer Science and Engineering, Mar Athanasius College of Engineering, Kothamangalam, India.
2Fathimathul Harshima P.T., Department of Computer Science and Engineering, Mar Athanasius College of Engineering, Kothamangalam, India.
3Prof. Aby Abahai T., Department of Computer Science and Engineering, Mar Athanasius College of Engineering, Kothamangalam, India.
Manuscript received on January 20, 2020. | Revised Manuscript received on February 05, 2020. | Manuscript published on February 29, 2020. | PP: 4160-4164 | Volume-9 Issue-3, February 2020. | Retrieval Number: C6627029320/2020©BEIESP | DOI: 10.35940/ijeat.C6627.029320
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Driver sleepiness is one in all the most important causes of most of the accidents within the world. Detecting the driver’s eye weariness is the simplest way for detecting the somnolence of the driver. The prevailing systems in the literature cannot discover sleepiness in folks having lagophthalmos (condition that forestalls eyes from closing utterly) and monocular vision (person with one-eyed or sight loss in one eye). To solve this downside a Driver State Alert Control system is projected which makes the use of head-shoulder inclination, face detection, eye detection, emotion recognition, eye openness estimation and blink counts for detecting the sleepiness and collision liability associated with robust emotional factors. The projected framework endlessly analyses the head shoulder inclination and facial lineaments of the driver to alert the driver by activating the alarm once he/she is drowsy or showing emotion unstable to drive. The typical separation between upper eyelid and lower eyelid of adults suffering from lagophthalmos is about 1-5mm. The EAR (Eye Aspect Ratio) is calculated supporting this separately for each eye. Thus, the proposed technique can be used for folks with lagophthalmos and monocular vision. Also, the entities don’t solely rely on blink count to ascertain the sleepiness, collision risks related to robust emotional factors are considered too. The projected system which is enforced with one camera view on open CV and raspberry pi setting illustrates the systems good efficiency in particulars of authentic sleep identification results and thus reduces road mishaps. It is easy to place in any type of vehicle and price effective too.
Keywords: Lagophthalmos, EAR, Open CV, Raspberry pi.