A Template Matching and Ellipse Modeling Approach to Detecting Lane Markers - Advanced Concepts for Intelligent Vision Systems - page 188

Table 1. Accuracy of the proposed lane detection system

Left Lane Right Lane
Correct Avg. E(f) ft. Correct Avg. E(f) ft.
Clip 1 96.31 % 0.018 97.21 % 0.035

Clip 2 95.92 % 0.033 96.94 % 0.024

Clip 3 94.26 % 0.012 95.50 % 0.020

Clip 4 82.08 % 0.012 94.02 % 0.019

Clip 5 100 % 0 99.48 % 0.012

Clip 6 95.78 % 0.014 99.53 % 0.020

Clip 7 100 % 0 100 % 0

Clip 8 73.28 % 0.019 96.50 % 0.032

9Con lu on

A new lane detection system is presented in this paper. At first, the input image
undergoes a geometric transformation followed by a color space conversion. Then
the procedures for detecting lane markers using template matching, morphology,
and elliptical modeling are explained. Kalman filtering and RANSAC used in
tracking and outlier elimination greatly helps in handling lane marker extraction
on curves. Finally, a new technique to calculate lane detection errors is also
introduced. Despite the presence of scattered shadows, illumination changes,
surface irregularities, and vehicles in neighboring lanes, our proposed system
showed very good performance which is portrayed by the quantitative results in
Table 1.

10 Future Work

The application of constraints on the width between the detected lane markers is
being explored. This should help prevent the sporadic oscillations of lane marker
estimates.

References

1. Lim, K., Seng, K., Ang, L., Chin, S.: Lane Detection and Kalman-Based Linear-
Parabolic Lane Tracking. In: 2009 International Conference on Intelligent Human-
Machine Systems and Cybernetics, pp. 351-354. IEEE, Los Alamitos (2009)
2. Lee, J.-W., Cho, J.-S.: Effective lane detection and tracking method using statisti-
cal modeling of color and lane edge-orientation. In: Fourth International Conference
on Computer Sciences and Convergence Information Technology, ICCIT 2009, pp.
1586-1591 (2009)
3. Wu, H., Liu, Y., Rong, J.: A lane detection approach for driver assistance. In: Inter-
national Conference on Information Engineering and Computer Science, ICIECS
2009, pp. 1-4 (December 2009)