THE DESIGN OF LIDAR-BASED OBJECT DETECTION AND GRAPHICAL USER INTERFACE SUBSYSTEM ON DROWSY DRIVER DETECTION SYSTEM

• Main Author: Rosa Savera Putri, Teresia
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/41738
• Institution: Institut Teknologi Bandung

Drowsy driver can pose a risk of accidents, especially when the driver was driving in a highway. Based on this problem, a device that can detect and alert drowsy drivers based on the driver's face and driving behavior is needed. The system developed in this final project consists of 3 subsystems, face reader subsystems to evaluate the driver's eye condition, lane departure detection subsystem to detect changes in the position of the car to the lane, object detection subsystem to keep the car at a safe distence with the vehicle in front of it, and graphical user interface associated with presenting warnings to the driver. This final project will specifically discuss the development of the object detection and grapichal user interface subsystem in drowsy driver detection system. The object detection subsystem in this device is activated when the driver's face shows some signs of drowsiness to prevent the forward collision accident due to the driver’s drowsiness. The object detection subsystem was developed using the Lidarlite v3 sensor and the graphical user interface subsystem is developed using the C ++ based Qt platform. The test results show that the object detection subsystem is able to read distances up to 40 m with a maximum tolerance of 166 cm and the graphical user interface successfully presents a visual warning for all types of warnings. The graphical user interface subsystem also manages to save warning data into the database and display it to the users.