Development of a Drowse and Erratic Driving Detection Dashcam using Haar Cascade

This study aims to create and design a Smart Dashcam with a Drowsy and Erratic Driving Detector, incorporating features not commonly found in current dashcam models. The development of the system involves the algorithm design, Haar Cascade. This algorithm uses a total of 68 face landmarks that were used in detecting a person's face in real-time. This Smart Dashcam with Drowse and Erratic Driving Detector specifically detects the eyes and mouth of the driver even at night. A GPS module was also integrated into the system for the additional feature of the dashcam such as speed monitoring, location monitoring, and data transmission. The speed monitoring feature determined the erratic driving detection of the device, and all data gathered was sent to the designated cloud server such as the ThingSpeak. The Smart Dashcam with Drowse and Erratic Driving Detector successfully incorporated multiple innovative features. For drowsy driving detection, it utilized real-time facial recognition and showed good results in detecting drowsy driving even at night by utilizing a web camera. The GPS module demonstrated its capability to accurately detect the driver's current location using latitude and longitude coordinates. After all, it was concluded that the Smart Dashcam with Drowse and Erratic Driving Detector can be improved by customizing the triggering threshold for consistency and accuracy.