This Project is based on the fourth task of the Udacity Self-Driving Car Nanodegree program. The main goal of the project is to write a software pipeline to identify the lane boundaries in a video from a front-facing camera on a car.

It was implemented in Python with OpenCV library. The project repo.

The key idea of the realization is the use of virtual sensors with adaptive position and threshold which analyze the region of interest row by row. It also uses information from previous frames and different techniques of results filtering for smoothing. The code also estimates the radius of the road curvature and the position of the vehicle with respect to the lane centre.

Contents:

  1. Camera calibration
  2. Image warping
  3. Math help function: Equidistant
  4. Adaptive Virtual Sensors

Final project videos:

The full project code is available on my Github

Project posts

Title img
07 Jun 2017

Adaptive Virtual Sensors for Lane detection

Computer vision approach for road marking detection with adaptive thresholds and positions of virtual sensors.

6 mins read
Title img
07 May 2017

Bird's Eye View Transformation

Convert a normal image to a Bird's Eye view projection with OpenCV.

3 mins read
Title img
05 May 2017

Camera calibration with OpenCV

A simple way to calibrate an optical system with a chessboard pattern by means of the OpenCV to reduce distortion.

3 mins read
Title img
02 May 2017

Approximate equidistant curve for polynomials

An approximate approach to find an equidistant curve for a given polynomial.

13 mins read