Implementation of Lane detection in Autonomous vehicle using FPGA

Lane Detection and Tracking have become an active research area for the past two decades, primarily in driver assistance applications. It is also an essential method for imaginative, predictive, primarily based fully autonomous vehicle devices. The main motto is to avoid human error, increase safety, reduce road accidents, and thereby save lives. Intoday’s autonomous vehicles, a variety of Advanced Driving Assistance Systems (ADAS) are used. Lane departure warning system is important ADAS which alerts the driver when inadvertent lane switching occurs. First, the image dataset is taken, and then the grayscale conversion takes place. The Gaussian filter reduces the impact of image noise. The Canny edge detection method and Sobel edge detection are then used to detect edges. The Edge detection method that detects a broad and wide range of edges in the images by using a multi degree system of rules. By using Hough transform method to detect the lanes after the edge detection process was done. Finally, the algorithms are implemented in FPGA and stored in BRAM (Block Random Access Memory) which helps to store the large amount of data in FPGA (Field Programmable Gate Array) technology Zynq 7000 SoC (System-on-Chips) using a platform known as Vivado HLS (High-level synthesis). The algorithm in an autonomous vehicle is formed on computer vision technique to find the road lanes and locate the driving lane. The algorithmsare executed to perform specific tasks on images to achieve highperformance, efficiency, and execution speed in terms of lane detection progress has been obtained.