Abstract: To obatain more ideal edge image for lane mark idenfitication, a lane mark identification and track method based on entropy maximization edge extraction was proposed. This method combined both onedimensional and twodimensional histogram entropy maximization partition methods, and image window variation technology was employed to partition the lane mark edge of the road image, then the lane mark parameters were acquired by the statistical method, finally the lane mark realtime track was performed by estabishing trapezoid areas of interest. Experimental results show that the proposed method is characterized by realiability, good robustness and realtimeness.

Key words: transportation system engineering, onedimensional entropy, twodimensional entropy, edge extraction, lane mark identification, lane mark track