A lane and curve detection using novel pre-processing with OpenCV

• Main Authors: Yalabaka, Srikanth, Tejaswi, Aravelli, Nethaji, Acha, Prasad, Ch. Rajendra, Vamshi, Konne, Kumar, Naveen
• Format: Conference Proceeding
• Language: English
• Subjects: Automatic pilots; Automobile industry; Automobiles; Autonomous cars; Autonomous vehicles; Feature extraction; Navigation systems; Preprocessing; Roads; Robotics
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0198890

Autonomous robotics and vehicles have a wide range of uses in self-driving cars as automation and electrification progress. Everywhere self-driving vehicles are trending in the automobile industry. Autopilot vehicles are changing innovation in the automobile industry with the best technologies. An important part of autopilot vehicles is the navigation system. While making or using autopilot vehicles, safety is the most important criterion for everyone. Finding the lanes and the curves is crucial for self-driving cars to function properly. The car can be driven more safely with the use of lane and curve detection. In this paper, we introduced a novel preprocessing method openCV. In preprocessing stage, the HUV color transform is employed to extract the white features and add preliminary edge feature detection. Then, the ROI is selected using the proposed preprocessing technique. This new preprocessing method is used to detect the lane. The performance of proposed method is evaluated using the standard KITTI road database. The results obtained are superior to the existing preprocessing and ROI selection techniques. The proposed model finds the road’s perfect boundaries to monitor it as a lane occurs in real-time.