验证码识别一般过程是：



二值化图像—>杂点过滤—>字符位置搜索—>字符点阵数量统计—>与字符点阵模版匹配



就拿我常去的一个执业药师的培训网站练练手，因为他的验证码非常的规则：没有杂点、字符位置固定、取字模之后100 匹配。时间有限，忙里偷闲研究了一下。VB的代码基本上是以前就写过了，也参照了一份网上流行的验证码识别源程序。

-Code identifying the general process is: binary image->　noise filter->　character location search->　number of characters in dot matrix Statistics->　and the characters dot matrix template matching Take I often go to a licensed pharmacist of the training site and thought hand because he rules the verification code is: no noise, the character position is fixed, take the font to 100　after the match. Time is limited, sneak in a little research. VB code is basically written before, and also with reference to a popular online source verification code identification.