基于混合特征的车牌定位算法

车牌定位技术是汽车牌照自动识别和智能交通系统的用车牌的颜色、纹理和结构几何等多维特征,实现车牌定位.该算法利用车牌的彩色信息进行彩色分割,实现车牌图像的二值化,而后提取边缘增强,在此基础上利用数学形态学方法去噪并去除车牌边框,并利用车牌纹理特征利用投影实现车牌的最终定位.该算法克服了单一特征信息不完备引起的车牌定位误差,实验表明该方法具有较好的车牌定位效果.

License plate location based on mixed characteristics

The vehicle license plate location is one of the key technologies in the license plate automatic recognition and intelligence traffic system. An algorithm of license plate location based on mixed characteristics of license plate is proposed. It make use of the license plate's multi-dimensional characteristics such as color, texture characteristic, shape, carrying out a license plate fixed position. Firstly, the input picture is roughly segmented based on the color information. Then a vertical edge detection method is proposed in con- sideration of the character texture, and a morphology filter is applied to merge candidate regions and remove noise. Finally the projective algorithm is adopted to get the plate area in the image. This algorithm overcome the car license position error which caused by the sole characteristic information is not complete. The experiment indicated the method has the better license plate location result.