METHOD AND SYSTEM FOR REAL-TIME CONTINUOUS LANE MAPPING AND CLASSIFICATION

The present disclosure discloses systems and method for real-time continuous lane mapping. The technique comprising receiving an input image of a road captured by an image sensor mounted on a vehicle; extracting one or more feature vectors from the image; extracting lane mark coefficients and lane type features from the one or more extracted feature vectors; detecting a lane mark by computing the coefficients and applying a pre-learned value; comparing the lane type relating to right and/or left lane markings with a predefined lane class; classifying the left and/or right lane markings based on the comparison and applying a pre-learned value; and generating a lane map along with the lane markings. The step of the lane detection and the classification is performed simultaneously based on the one or more extracted feature vectors.