Lane Detection Algorithm Based on Reliable Lane Markings for Self Driving Vehicles
I B V S Shiva Sai1, K Sathish2, G Aurava3, S Mahboob4, K N V C Koushik5
1I B V S Shiva Sai, SRM Institute of Science and Technology, Kattankulathur (Tamil Nadu)-603203, India.
2K Sathish, SRM Institute of Science and Technology, Kattankulathur (Tamil Nadu)-603203, India.
3G Aurava, SRM Institute of Science and Technology, Kattankulathur (Tamil Nadu)-603203, India.
4S Mahboob, SRM Institute of Science and Technology, Kattankulathur (Tamil Nadu)-603203, India.
5K N V C Koushik, SRM Institute of Science and Technology, Kattankulathur (Tamil Nadu)-603203, India.
Manuscript received on May 05, 2019. | Revised Manuscript received on May 15, 2019. | Manuscript published on May 30, 2019. | PP: 5-9 | Volume-6 Issue-3, May 2019. | Retrieval Number: C2236046319/19©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Keeping the vehicle within the lane is an important aspect in the self driving vehicles. These lane lines are detected by using lane detection algorithms. Initially the self driving vehicle captures the footage of the road ahead of it using high resolution cameras mounted on the top of the car. Then the footage is divided into individual frames and the frames are processed further to identify the lane markings. Digital image processing technique is utilized in order to find ROI (Region of Interest) and to eliminate unnecessary noises and glares caused by the reflection of light. Then, the light intensity and width of lane markings are taken as input. An edge detection algorithm is applied to find the boundaries of objects within images. It works by detecting discontinuities in brightness followed by a line detection algorithm is applied on the edge detected image to construct the lines on which the edge point lies. Hough transform with some subsidiary conditions is suitable algorithm preferred. With this proposed model, the lane can be accurately detected in conditions of fluctuating, poor illumination and from interference from reflected light can be avoided effectively. The results obtained demonstrate the accuracy of the proposed method.
Keywords: Self Driving Vehicle, High Resolution Camera, Edge detection, Morphological filters, Hough Transform.