point:如果因为获取验证码的教务系统响应缓慢或者无法响应，正确率测试的地址会因为无法获取到验证码图片而产生504错误/响应过长，可以考虑把脚本下载到本地将目标url换成本校的教务系统验证码url再做测试
- PIN Identify by fangzheng.php 为主文件
- PIN Identify lib.php 为其函数库
- downloadImg.php 用于download验证码
- AddDictionary.php 用于添加字典
- zidian.sql 为字典，建一个叫'yzm'的数据库导入即可
(已恢复)正确率测试 地址
(已恢复)正确率批量测试 地址
批量测试设置了一次加载五个方正验证码，大概等待时间在17s左右，而且可能会因为请求不到验证码而中断
kuri-su.cc
测试页面出现错误的情况已经修复
不嫌弃的话给个star吧wwww
PS：目前认为运算效率的瓶颈在 similar_text() 这个核心函数上，函数的作用在于对比两个字符串的相似程度，但是其算法时间复杂度T(n) = O(n^3) ，所以准备用T(n) = O(m*n)动态规划实现一个相类似的函数去代替该函数，以提高性能，时间未定。。。
There are 2,300 records in the current dictionary，Recognition rate of about 85%，You can make sure that the request is sure to be correct once，The single recognition run time may be between 2.5-3s。
- PIN Identify by fangzheng.php is main file
- PIN Identify lib.php is function lib
- downloadImg.php Used to download the verification code
- AddDictionary.php Used to add a dictionary
- zidian.sql Is a dictionary，Build a database called 'yzm' and imported
to Accuracy test
to Batch test accuracy
Batch test set a load of five Founder verification code, probably waiting for about 17s time, and may not be required because the request code is interrupted
kuri-su.cc
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent