Route perception of autonomous vehicle using feature extraction and KALMAN filter

A critical and essential component of modern autonomous driving systems is Lane detection as it enables vehicles to navigate safely and accurately on roads. The capability of recognizing and tracking lane markings is very crucial to maintain proper lane positioning, avoiding collisions, and ensuring the safety of both passengers and other road users. However, lane detection can be challenging, particularly in complex environments with adverse weather conditions. This paper introduces a novel algorithm for lane recognition aimed at tackling the aforementioned challenges. and enhances the accuracy and stability of lane recognition, even in challenging conditions like complex and foggy environments. The proposed method involves a series of pre-processing techniques such as grayscale, blurry, and gradient calculation, followed by the canny operator for edge detection. The proposed lane recognition approach involves the application of the Canny algorithm for edge detection to accurately detect lane markings, followed by the utilization of the Hough transform to extract the extreme edges of the lane lines. Additionally, to enhance the accuracy and stability of lane detection in challenging environments such as low light or heavy rain, the Kalman filter is employed to track the boundaries of each lane marking extracted via the Hough transform. The suggested algorithm has the potential to enhance the chances of identifying lane markings even in difficult circumstances. This improvement can play a crucial role in ensuring the safe navigation of autonomous vehicles.