A study on real time integrated lane detection and vehicle tracking method with side-mirror cameras

Estimation of the Lateral Distance between Vehicle and Lanes Using Convolutional Neural Network and Vehicle Dynamics

“…An alternative method is introduced to accelerate the selection of texture blocks. Before the image is synthesized, feature information on the edge of texture blocks is calculated by (2) and is stored into the database. Smaller cost means higher similarity.…”

“…Thanks to the development of big data and computer vision and other advanced technology [1][2][3], modern cars are equipped with a large number of Advanced Driver Assistant Systems (ADASs), such as Lane Departure Warning (LDW) and Lane Keeping Assist Systems (LKAS), which are designed to perform the functions of lane departure warning and keeping [4]. However, before vehicles with these safety applications are launched on the market, numerous road tests ought to be carried out.…”

Lateral distance detection model based on convolutional neural network

“…Therefore, the complexity of data processing is reduced and robustness of lateral distance is improved. Since the ROI is limited to the limited area on both sides of a vehicle, a lower resolution camera can be used [30].…”

Fast and robust approaches for lane detection using multi‐camera fusion in complex scenes

“…Also, vision sensors are cost‐effective, affordable, and easily integrated. However, the existing vision‐based lane detection methods either use a single front‐view camera [14, 15], or stereo camera [5], or two side‐mirror cameras [16] for road image acquisition. Although it can meet the requirements of lane detection under the common smooth road sections, the method could not get a good detection result in the traffic jam with heavy traffic or serious occluded situation.…”