基于数学形态学和颜色特征的车牌定位方法

为了解决单一车牌定位算法在复杂背景中定位效果不理想的现状，提出一种数学 形态学和颜色特征相结合的算法对车牌进行定位。首先利用最大类间方差法(Ostu 算法)找到一 个最佳的阈值，根据所得阈值把得到的灰度图像二值化，然后采用一种改进的数学形态学算法 对图像进行边缘检测，最终结合数学形态学和车牌颜色特征进行准确定位。实验表明该算法明 显优于传统或单一的车牌定位方法，定位准确率高，对背景限制少，应用范围广。

Method of License Plate Location Based on Mathematical Morphologyand Color Characteristics

In order to solve the problem that the situation of single license plate localization algorithm in complex background is not ideal. A license plate location algorithm based on mathematical morphology and color characteristics is proposed. Firstly, the new method uses Ostu algorithm to select an optimal threshold, and according to the threshold carries on binary processing, then by using a new method of mathematical morphology for image edge detection. Finally combined mathematical morphology with color characteristic to locate license plate. Experiments show that this new method with higher accuracy and fewer restrictions on background is superior to traditional or a single method and can be widely used.