从加强网站安全性的角度,研究认证码字符识别的技术要点.首先,提出预处理算法去除图像中的干扰因素,分割出字符;其次,采用分类器技术对预处理后得到的字符进行识别.实验指出有效的预处理算法和选择合适分类器对于快速准确识别认证码起着至关重要的作用,证实了目前一些认证码在保障网站安全性的同时,尚存在一些漏洞,需要继续完善.
Recognition of authentication code is investigated from the perspective of strengthening website security. A preprocessing algorithm is first developed to remove the disturbance factors in the image and obtain the characters, followed by using some classifiers for recognition purposes. The experimental results suggest that the developed algorithm and selected classifiers are the most critical part in the recognition process. Some loopholes are still identified in the authentication codes in the website protection phase, and hence further investigation into this issue is needed.