车牌图像预处理及识别算法研究

实现了车辆牌照自动识别系统中的一套预处理方法,主要包括车辆图像的灰度化,灰度级调整,消除噪声,图像二值化等。通过灰度级跳动来实现车牌定位及分割,最后利用模板匹配自动识别车辆牌照信息并输出结果。本文采用的灰度变化及灰度级调整可有效的消除一些外部环境对车辆牌照的影响,在一定程度上提高了对车辆牌照的定位与分割的准确性,可对光线较昏暗,车牌较模糊的图像也可较准确的进行定位和分割。

The Preprocessing and Recognition Method for License Plate

A set of pretreatment methods of the vehicle license plate recognition system are provided in this paper,including grayscale images,gray scale adjustment,eliminate noise,image thresholded and so on.The license plate positioning and segmentation are achieved through gray-scale pulsation.Vehicle license information is automatic identified by template matching and output the information.In this paper,the impact of the external environment on vehicle license can be effectively eliminated by the gray scale change and gray-scale adjustment,to some extent,the position of the vehicle license and segmentation accuracy can be improved.This method is appropriate for the dim light,blurred images of license plate.