A robust and real-time image based lane departure warning system

Lane Departure Assessment via Enhanced Single Lane-Marking

“…Finally, only 604 in 1546 frames include two lane-markings in our dataset. The best performance on our dataset among all six is [15], which failed to reach the 90% correct warning rate. The main reason is that these algorithms need two lane-markings while the angles between the two detected lane-marking lines might change drastically with the deviation angles of the camera.…”

“…A lane departure identification method used three lane-related parameters, including the Euclidean distances between every two points of the Hough origin H o , the midpoints mp 1 and mp 2 of the identified left and right lane-markings to identify the state of departure [11][12][13]. Besides, algorithms judging the (ρ, θ) patterns or just one of the detected left, and right lane-markings determined the left or right lane departure situation [14][15][16][17][18][19][20][21][22][23][24]. The recent study conducted by Lin et al determines lane departure also by the information of the detected lane-markings only, and it uses a state machine to recognize the "left," "right," and "normal" status, which can reduce the false alarms when the lane-marking is blocked by obstacles [3].…”

International Journal of Electronics and Telecommunications

“…Consider the image is shown in Figure .2, with set sample of detected edge point co-ordinates as (xi, yj) = (6,5), (7,6), (7,16), (8,15), (9,14), (10,13), (13,11), (14,10), (15,9), (16,8) and (17,7).…”

“…In another recommendation, 40% of image from bottom has been considered as ROI [1]. Another proposal has introduced an intelligent trapezoidal ROI [8]. The ROI selection, line detection procedure has been applied independently on the first frame and carried forward to the succeeding video frames [11].…”