Forward-Looking Lane Marking Detection and Tracking

LMDT IP core implements an algorithm for lane markings extraction and tracking using a camera installed in front of a vehicle.

It can be used as base component of a forward-looking Lane Departure Warning (LDW) system whose objective is to alert the driver when the vehicle unintentionally departs from its road lane. Indeed, that system requires to identify and track the markings corresponding to the lane boundaries and to locate the relative host vehicles position.

LMDT aims to a long range lane detection using a curvilinear lane model.

LMDT structure

The module is composed by a pipeline of functional blocks implementing:

Pre-processing: extraction of the lane marking candidate points
Model- fitting: robust fitting of the mathematical lane model to the set of points
Model tracking: tracking of the lane model parameters in time

The model-fitting is a variant of RANSAC method (RANdom Sample Consensus), while the model-tracking is based on Kalman filters.

Key Features

Fundamental building block for FPGA based automotive Forward Looking Lane Departure Warning System
Supports Spartan®-3ADSP and Spartan®-6 Xilinx® FPGAs
Adapts to shadows and light changes
RANSAC based model fitting
Implements the most computing demanding tasks in programmable logic
Data Rate 640x480@60Hz
Fully embedded into Xilinx XPS and the EDK

Purchase

Reference design for XtremeDSP Video Starter Kit (Spartan-3A DSP Edition) is available. For ordering information please contact This email address is being protected from spambots. You need JavaScript enabled to view it. .

Publications

For technical details please refer to the following publications:

"Building Automotive Driver Assistance System Algorithms with Xilinx FPGA platforms" - Xcell Journal, fourth quarter 2008 - Download PDF
"A real-time versatile roadway path extraction and tracking on an FPGA platform" - Computer Vision and Image Understanding, Volume 114, Issue 11, November 2010, Pages 1164-1179. Special issue on Embedded Vision