一种对粘连英文字符串的快速切分算法研究

摘要

Under themature conditions of English character recognition,the string segmentation becomes the key step. Especially adhesion English string model is very complex,which is difficult to use the traditional image processing method for accurate segmentation. To solve this problem,present a simple and fast segmentation algorithm. It extracts upper and lower concave area through computing the image’ s background. Then it determines the true segmentation region using adjacent matching principle and the minimum area principle. Finally, it can extract segmentation cutting line successfully. The experimental results show that the algorithm can obtain better results on English string,especially on segmentation of adhesion on English string,which is satisfactory.%就现有英文字符识别技术成熟的条件下，字符串的切分成为非常关键的环节。其中粘连英文字符串的模式尤为复杂，难以用传统的图像处理方法进行准确地分割。针对该问题，文中提出了一种简单快速的切分算法。该算法通过计算图像的背景域，从而提取出上下凹区域，再采用相邻匹配原则和最小面积选择原则最终确定切分域，从而成功提取出切分线进行准确切分。实验结果表明，该算法对英文字符串切分效果良好，尤其对粘连英文字符串也可以得到较好的分割结果，令人满意。