Lane Lines Detection under Complex Environment by Fusion of Detection and Prediction Models

Abstract

The lane lines’ length, width, and direction are very regular, serialized, and structurally associated, which are not easily affected by the environment. To enhance lane detection in a complicated environment, an approach combines visual information with the spatial distribution. Firstly, the grid density of the target detection algorithm YOLOv3 (you only look once V3) is improved from S×S to S×2S, aiming at the particular points in the bird’s-eye view where the lane lines had different densities in the horizontal and vertical directions. The obtained YOLOv3 (S×2S) is more suitable for detecting objects with small and large aspect ratios. It also identifies image features along with balances the detection speed and accuracy. Secondly, based on a bi-directional gated recurrent unit (BGRU), a new lane line prediction model BGRU-Lane (BGRU-L) based on the distribution of lane lines is proposed using the characteristic of lane line serialization and structural correlation. Finally, Dempster-Shafer (D-S) algorithm based on confidence was used to integrate the results of YOLOv3 (S×2S) and BGRU-L to improve the lane line detection ability under complex environments. The experiment was carried out on the Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI) dataset, while Euro Truck Simulator 2 (ETS2) is used as a supplement dataset. After fusing YOLOv3 (S×2S) and BGRU-L models in the D-S model, the detection results have high accuracy in a complex environment by 90.28 mAP. The detection speed is 40.20fps, which enables real-time detection.

Object Detection in Autonomous Vehicles under Adverse Weather: A Review of Traditional and Deep Learning Approaches

“…In [169], the authors modified YOLOv3 and presented a brand-new BGRU-Lane (BGRU-L) model; the method integrates spatial distribution with visual data. High accuracy (90.28 mAP) and real-time detection speed (40.20 fps) are achieved in difficult settings through integration utilizing the Dempster-Shafer algorithm, as demonstrated by datasets from KIT, Toyota Technological Institute, and Euro Truck Simulator 2.…”

Fast and accurate object detector for autonomous driving based on improved YOLOv5

“…They constructed a two-stage network based on the YOLOv3, presented a way for the automatic generation of the lane label images in simple scenarios, and used an adaptive edge detection method based on the Canny operator to relocate the lane recognized by the firststage algorithm. To improve lane detection performance in a complicated environment, Haris et al 21 proposed an approach combining visual information and spatial distribution, which improves the grid density of the object detection algorithm YOLOv3 and presented a new lane line prediction model BGRU-Lane. To solve the problems of low detection accuracy and poor real-time performance of traditional methods, Huu et al 22 advised a lane and object detection algorithm, which improves the quality of the distorting image caused by the camera, implements the sliding window to determine pixels of each lane, and utilizes YOLO algorithm to identify lanes and obstacles (Table 1).…”

Based on Haar-like feature and improved YOLOv4 navigation line detection algorithm in complex environment

“…But too much network fusion increases the detection time and cannot meet the real-time requirements. Haris, M [8] et al proposed a lane line prediction model BGRU-Lane based on lane line distribution, and the dempster-Shafer algorithm was used to integrate the results of BGRU-L and Improved YOLOv3 to improve the lane line detection ability under complex environments. YuCheng Fan [4] et al combined lidar detection in unmanned driving technology with YOLOv4, and proposed the LS-R-YOLOv4 algorithm, which improved the accuracy of target detection in unmanned driving.…”