基于数学形态学车牌自动定位和倾斜校正

针对车牌识别系统中,由于光照变化、背景复杂、车牌褪色或相对摄像机位置不同带来的图像倾斜等原因,导致车辆图像质量不高,影响车牌定位和识别。基于优化组合数学形态学基本运算,结合灰度信息离散化对车牌初定位,再利用先验知识,对车牌进行精确定位,最后采用两点法对倾斜的车牌图像进行检测和校正。实验结果表明,得出的定位和倾斜校正方法简单、快速,且具有较好的稳定性。

License Plate Automatic Location and Tilt Correction Based on Mathematical Morphology

In the license plate recognition system, factors like poor illumination, complex background and the tilt of the license plate will cause the inaccurate location and recognition of the license plate. The paper brings forward a method to accurately locate license plate by optimizing basic operation combination of mathematical morphology, combined with discrete gray information and prior knowledge. It corrected tilt of license plate image by two-points method. The experimental results show that, the orientation and tilt correction method is simple, rapid, and can well eliminate interference with high accuracy and good stability.