摘要
传统论文查重系统往往关注论文的完整性和真实性,但互联网新兴媒体发达时代,新的挑战层出不穷。毕业论文中,特别是平时的课程论文,图像窃取现象十分严重,而目前的查重工具往往不能实现图片查重,需要一个以人工智能图片查重算法为基础的图片匹配度查重系统。SIFT算法以匹配时间短、准确性高、实时性强著称,得到了广泛应用。基于此,总结了目前SIFT及其他相关图片匹配度算法,以平时学生毕业论文、课程中的图片作为实验对象,在充分考虑实践各方面可能性的基础上进行实验检测,利用KNN匹配优化算法,得出最终合适的图片匹配度算法,同时,分析实践中遇到的关键问题,提出改进建议。
Traditional paper duplication checking system often pays attention to the integrity and authenticity of papers,but in the era of Internet emerging media,new challenges emerge endlessly.In the graduation thesis,especially in the usual course papers,the phenomenon of image theft is very serious,but the current duplication checking tools often can not achieve image duplication checking,which requires a duplication checking system based on artificial intelligence image duplication checking algorithm.SIFT algorithm is famous for its short matching time,high accuracy and real-time performance,and has been widely used.Based on this,this paper summarizes the current SIFT and other related image matching algorithms.Taking the pictures in students’graduation thesis and courses as the experimental objects,and taking full account of the possibilities of various aspects of practice,the experimental detection is carried out.The KNN matching optimization algorithm is used to obtain the final appropriate image matching algorithm.At the same time,the paper points out that the image matching algorithm is suitable.The key problems encountered in practice are analyzed and suggestions for improvement are put forward.
作者
丁一
Ding Yi(Wuxi City College of Vocational Technology,Wuxi Jiangsu 214153,China)
出处
《信息与电脑》
2019年第16期37-40,共4页
Information & Computer
基金
2019年江苏省职业教育教学改革研究课题“新一代信息技术支持下构建基于过程的课程考核体系的实践研究”（项目编号：ZYB345）
