An Optimization Algorithm of Vehicle License Plate Correction Based on Minimum Projection Distance

This paper presents a method of vehicle license plate correction based on minimum projection distance of characters area. The method shows that the projection distance to vertical coordinate axis of character area is is minimum when the vehicle license plate is corrected. Then computing the minimum vertical projection distance can capture the tilt angle of slant image. And similarly the shear angle can be computed from horizontal minimum projection distance by shearing vehicle license plate. At last bilinear interpolation achieve the vehicle license plate correction image. The method of binary search and virtual rotation are used to optimize the correction algorithm. The correction method with good anti-interference doesn´t depend on the frame-line of the vehicle license plate. Actual Chinese vehicle images taken under various conditions from traffic stations are used in the experiments to evaluate the robustness, the flexibility and effectiveness.