Vision based lane detection in autonomous vehicle

A novel method for finding and tracking road lanes for vision-guided autonomous vehicle navigation is present in this paper. First an inverse perspective mapping is applied in order to remove the perspective from the camera observed image. Then the edges of the road lanes are detected from the inverse perspective mapping images. A particle filtering algorithm is used which allows us to compute the likelihood between all the particles with the edge images, then the three parameters of the real state of lanes can be estimated. The new lane detection method described in this paper has been tested in real road images and the result is robust and reliable.