真迹防伪标签自动识别算法研究

发布时间：2018-06-22 10:50

本文选题：防伪 + 特征提取　；参考：《海南大学》2017年硕士论文

【摘要】：假冒伪劣产品危害消费者的健康,损害合法经营者的利益,影响社会经济的发展,防伪技术的相关研究正是为了打击造假者,维护市场经济秩序。由于防伪技术具有一定的时效性,因而越是新型的复杂的防伪技术越能起到防伪的效用。因此,本文选取了一种新型的防伪技术——真迹防伪技术作为研究对象。真迹防伪的防伪标签不易复制仿造,与互联网结合进行自动识别,既能方便消费者的使用,解决肉眼识别的不便,又能对真迹防伪进行一定程度的推广。因此本文做了如下工作:(1)针对真迹防伪技术这一新型的防伪技术,对其防伪标签现在不能进行自动识别的问题进行研究和解决。(2)研究了基于DFT、基于DCT和基于DWT-DCT的三种图像特征提取方法,并验证其可行性和可靠性;(3)根据所研究的的三种图像特征提取方法,提出了三种对真迹防伪标签自动识别的算法,并通过实验验证三种自动识别算法的抗攻击能力、识别时间等识别性能。(4)对以上三种真迹防伪标签自动识别算法的各项识别性能进行比较分析,得出一种适合真迹防伪标签的自动识别算法。并对真迹防伪标签自动识别算法的未来研究提出了设想。
[Abstract]:Fake and inferior products are harmful to the health of consumers, to the interests of legal operators, and to the development of social economy. The research on anti-counterfeiting technology is aimed at cracking down on counterfeiters and maintaining the order of market economy. Because the anti-counterfeiting technology has certain timeliness, the more complex anti-counterfeiting technology can play the role of anti-counterfeiting. Therefore, this paper selects a new anti-counterfeiting technology-genuine anti-counterfeiting technology as the research object. The anti-counterfeiting label is not easy to copy and copy, and can be recognized automatically by combining with the Internet. It can not only facilitate the use of consumers, solve the inconvenience of naked eye recognition, but also promote the anti-counterfeiting of authenticity to a certain extent. Therefore, the following work has been done in this paper: (1) aiming at the new anti-counterfeiting technology, The problem that the anti-counterfeiting label can not be automatically recognized is studied and solved. (2) three image feature extraction methods based on DFT, DCT and DWT-DCT are studied. The feasibility and reliability of the algorithm are verified. (3) according to the three image feature extraction methods studied, three algorithms for automatic recognition of genuine anti-counterfeiting labels are proposed, and the anti-attack ability of the three automatic recognition algorithms is verified by experiments. (4) comparing and analyzing the recognition performance of the above three automatic identification algorithms, an automatic recognition algorithm suitable for genuine anti-counterfeiting label is obtained. The future research of automatic recognition algorithm for genuine anti-counterfeiting label is proposed.
【学位授予单位】：海南大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【相似文献】