摘要
通过对车牌原始图像进行水平位置和垂直位置定位算法优化,提高超分辨率识别效果,实现了支持向量机的车牌识别目的;通过平滑处理对图像进行分割优化,保证图像噪音点能够在不同的频域产生高频分量,保证能够在图像处理过程中建立平滑处理方程,使图像处于平滑的状态。按照边缘提取控制方法实现超分辨率优化,达到了分辨率算法设计提升车牌识别性能的效果。
Through the optimization algorithm of horizontal and vertical position to locate the li cense plate images, the effect of super resolution was improved and the licenseplate recognition support vector machine was optimized. Smoothing segmentation of images was adopted to ensure that image noise points at different frequency domain can generate high frequency components. A smoothing equation is established for image processing so that the image is always in a smoothed state. Superresolution optimization is realized in accordance with the edge detection control method, and a resolution algorithm is designed to enhance the recognition performance.
出处
《淮海工学院学报（自然科学版）》
CAS
2013年第2期29-31,共3页
Journal of Huaihai Institute of Technology:Natural Sciences Edition
关键词
车牌识别
超分辨率
灰度
license plate recognition
super resolution
gray scale
