Thesis (M.Sc.) - Cairo University - Faculty of Engineering - Department of Mechanical Design and Production

This thesis presents a comprehensive mathematical modeling and simulation for the trajectory of a vision-based autonomous vehicle during moving between lane lines of the structured road. In addition, demonstration building, implementing, and developing a trajectory tracking control system based on computer vision for autonomous cars. The simulation accomplished by using MATLAB/Simulink software. This simulation mimics the existence of an actual digital camera by using a novel 3D-vision block to simulate the actual images that assumed to be provided by a digital camera connected to an embedded computer. The 3D-vision block uses mathematical equations, execution sequence and logical conditions to create a virtual captured image. So, this virtual image is then used to detect the lane in the front of the vehicle depending on the virtual camera position and its parameters. Inside simulation environment that based on the kinematic model of the vehicle and vision model, the controller is designed in the simulation and is coded in the embedded computerwith the optimized control gains. The implementation presents a system includes a single digital camera, an embedded computer(Raspberry Pi 2), and a microcontroller boardto produce an autonomous car to be able to track current road lane, where the digital camera is mounted at the top of the vehicle along its longitudinal axis. The real-time captured images are processed using Python code with OpenCV library over Linux operating system to obtain geometrical data of road lane. From this data, the observable errors can be determined. Finally, a steering controller utilizes these errors in control law that designed in the simulation with tuning in control gains to compute the steering command. The embedded computer then paths this command to Arduino microcontroller board to adjust the steering servomotor. During this work, a set of autonomous driving experiments is performed. Several evaluations scenarios are shown and discussed about lane detection

Issued also as CD