An Improved FAST Corner Detection Algorithm for Real-Time Lane Lines Detection on Road Surface

Detecting near lane lines presents challenges due to the influence of large noise blocks on the road surface. This study proposes an improved algorithm for realtime lane lines detection to counter this problem. The algorithm is carried out in three key steps: First, a minimum filter technique is performed to remove salt noise within the image. Subsequently, a modified Features from Accelerated Segment Test (FAST) corner detection algorithm is utilized to further filter and eliminate the block noise. Finally, Sobel edge detection algorithm is employed to achieve precise edge detection results for the lane lines. The proposed method has been deployed on a Field-Programmable Gate Array (FPGA), with comparative tests illustrating both its feasibility and effectiveness in practical application.