Smart car visual servo control based on OpenCV and Kinect (Course project)

Problem Description:

It was the final project of the course AU326 Fudamentals of Digital Image Processing. Each team was assigned with one small car with Bluetooth device, one wireless camera and one Kinect sensor. We completed two tasks: (1) extract the road information including the lane shape and the traffic signs using OpenCV (2) design a controller to drive the vehicle along the path and respond to the traffic signs. (3) develop a visual driving interface for human drivers to interact with the vehicle via Kinect.

My Contributions:

(1) Repackaged the car control instructions into a new c++ class, making it convenient for higher-level controller design.

(2) Developed the perception algorithm that extracted the lanes.

(3) Collaborating with Shaobo, we developed the driving interface shown by Fig 2 and 3.