位置:成果数据库 > 期刊 > 期刊详情页
基于RGB颜色模型及图像二值化的红晕印鉴提取
  • ISSN号:1006-9798
  • 期刊名称:《青岛大学学报:工程技术版》
  • 时间:0
  • 分类:TP391.41[自动化与计算机技术—计算机应用技术;自动化与计算机技术—计算机科学与技术]
  • 作者机构:青岛大学电子信息学院,山东青岛266071
  • 相关基金:国家自然科学基金资助项目(61272077)
中文摘要:

针对票据认证识别中印鉴难以准确提取以及放置时间过长的手工印鉴周围出现红晕而失效 的问题,本文提出一种带红晕印鉴提取的技术方法.先将图像进行二值化,采用RGB颜色模型提 取印鉴,并通过基于RGB颜色模型的图像提取方法与图像的二值化相结合的方法,找到去除红晕 等噪声印鉴的位置信息,同时根据找到的位置信息进行印鉴还原实验.实验结果表明,带红晕印鉴 提取的技术方法能有效去除红晕等噪声,准确提取票据上的印鉴.该方法较好地解决了带红晕等 噪声的印鉴提取难题,对印鉴识别技术的研究具有一定的实际意义.

英文摘要:

The problem about accurate extraction of seal hasn?t been solved well in the recognition of notes. The general extraction’s method based on RGB color model is very effective for extraction of electronic seal. But the method is very difficult for extraction of the manual seal especially extraction of the seal that arises blush, because of placing for long time. In view of the problem about seal extraction in note identification and the problem that it is invalid to extracting manual seal with blush, this paper proposed an extraction’s method about seal with blush as follow: firstly, image must be made binary processing. Then, seal ispreliminary ex-tracted by extraction of RGB color model. Next, using the method of the combine the method based on extrac-tion of RGB color model and image binarization finds locational information of seal of removing blush. Mean-while, the seal image is restored by the locational information. Experiments show that extraction ?s technology of seal with blush can effectively remove noise,like blush and accurately extracts seal of notes. This method is a better method for extracting seal with noise, like blush, and has practical significance for the study about seal identification.

同期刊论文项目
同项目期刊论文
期刊信息
  • 《青岛大学学报:工程技术版》
  • 中国科技核心期刊
  • 主管单位:山东省教育厅
  • 主办单位:青岛大学
  • 主编:邵峰晶
  • 地址:青岛市宁夏路308号
  • 邮编:266071
  • 邮箱:qdxbgc@126.com
  • 电话:0532-85953597
  • 国际标准刊号:ISSN:1006-9798
  • 国内统一刊号:ISSN:37-1268/TS
  • 邮发代号:
  • 获奖情况:
  • 国内外数据库收录:
  • 被引量:3616