L*a*b* color model based road lane detection in autonomous vehicles

Abstract

Autonomous vehicles, as a main part of Intelligent Transportation Systems (ITS), will have great impact on transportation in near future. They could navigate autonomously in specific areas or highways and city streets using maps, GPS, video sensors and so on. To navigate autonomously or follow a road, intelligent vehicles need to detect lanes. This paper presents a method for lane detection in image sequences of a camera on top of a robotic vehicle. The main idea is to find the road area using the L*a*b* color space in consecutive frames. Subsequently, by applying this model in road area and equalization of histogram and calculation of gradient image using Sobel operator, the parameters of the lane can be calculated using a Hough transform. The proposed method is tested under various illumination conditions and experimental results indicate the good performance of the proposed method.

Aircraft Navigation on Taxiways

“…When working on line detection, the reference color can easily vary from one location to another, or from the scene lighting. Some authors ( [8], [4]) propose to work in other color spaces, such as HSV, Lab or YCbCr, that have the particularity to differentiate lightness information from chromatic information. Other authors ( [12] and [5]) try to use a local adaptative threshold in order to be robust to shadowing or lightness variation.…”

“…In the case of the contrast between the line marking and the road, the ground lines are designed to have more reflectivity than the road. That is why most of the studies ( [14], [7], [2], [4]) try to detect line markings for lane detection applications, either by using color models, contrast or orientation of surfaces. The low computational complexity of those algorithms is an advantage but they are sensitive to shadows or occlusions.…”

Colorimetric Space Study: Application for Line Detection on Airport Areas

“…As such, it is widely present in the automotive field literature (Narote et al, 2018). A good number of studies use a color and/or contrast prior for line detection, where the color space used for this application varies, such as L*a*b* (or CIELAB) in (Kazemi and Baleghi, 2017), HSI in (Sun et al, 2006) or HSV in (Lipski et al, 2008) and (Mammeri et al, 2016). In the automotive field, markings are most of the time white, but their color could vary following the type of road or country.…”