摘要
该文提出一种基于小波变换和数学形态学的视频图像文本定位算法。该算法首先对输入的灰度图像进行小波变换．在对垂直方向高频分量图进行二值化、形态学操作和连通分量分析以及先验知识的限制后，定位出文本区域。
An algorithm on text location are proposed on the basis of wavelet transform and mathematical morphology in video images. The algorithm first input gray image wavelet transform, text area can be located after binarization, morphological operations and connected component analysis and a priori knowledge of the restrictions in the vertical high-frequency component images.
出处
《电脑知识与技术》
2009年第4期2720-2721,2728,共3页
Computer Knowledge and Technology
关键词
小波变换
数学形态学
wavelet transform
mathematical morphologyt
