Manuscript received on July 11, 2020. | Revised Manuscript received on July 15, 2020. | Manuscript published on July 30, 2020. | PP: 103-106 | Volume-9 Issue-2, July 2020. | Retrieval Number: B3156079220/2020©BEIESP | DOI: 10.35940/ijrte.B3156.079220
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: This paper describes an inexpensive distracted driver detection device built using a Raspberry PI3, a video camera, and python code. Distracted and drowsy driving are two of the leading causes of automobile accidents in the United States, and this inexpensive standalone can help prevent those deaths. live video feed of the driver’s face is read in and through the facial landmark detector within DLIB, the co-ordinates of the eyes and mouth are extracted in real time. The ratio of the distance between the eyelids in the vertical direction to the horizontal direction defined as Eye Open Ratio is used to evaluate if the eye is open or closed and a similar ratio on the mouth(Mouth Open Ratio) is used to identify if it is open or closed. By looking at the eye and mouth open ratio for several consecutive frames, it is determined with >95% accuracy whether the driver is drowsy, distracted, or yawning. If any of these behaviors are noted, the device will prompt an audible warning to encourage the driver to focus on the road. The prototype was tested under a variety of lighting conditions from dark to bright light and on different subjects with and without glasses. This test data was used to determine the threshold for when the eye or mouth is determines open or closed. Additionally, the prototype’s settings are customizable for a primary driver to further improve the accuracy. The device connects to a smart phone and sends information with the time stamp of the distracted driver incident. This device can be used to prevent distracted or drowsy driving-related deaths, and is an inexpensive attachment that can easily be fitted into a preexisting vehicle.
Keywords: Raspberry Pi, Python, Night vision camera. Distracted driving.