摘要
针对女书字符的结构特点,结合LLT局部阈值算法的基本思想,通过在局部窗口内引入基于最大类间方差的OTSU算法,并设计相应的阈值修正参数,给出了一种基于局部自适应的女书字符图像分割方法,从而实现了局部阈值的自动获取.通过与OTSU算法相比,该算法不仅具有较强的抗噪能力,而且较好克服了女书古籍文献中常见的光照不均匀性问题.实验结果进一步验证了该算法的适应性和鲁棒性.
For the structural characteristics of NvShu characters ,by combining the basic idea in LLT local threshold algorithm and introducing OTSU algorithm based on maximal between-class variance into local windows , and designing the corresponding modify parameters for the threshold , an image segmentation method for NvShu characters image based on adaptive logical algorithm was presented in this paper ,so that the automatic acquisition of local threshold was achieved .Compared with OTSU algorithm ,the proposed algorithm not only has strong anti-noise capability ,but better overcome the common problem with uneven illumination in NvShu ancient literature . Experimental results demonstrated its feasibility and robustness .
出处
《微电子学与计算机》
CSCD
北大核心
2014年第3期142-145,149,共5页
Microelectronics & Computer
基金
国家自然科学基金(60975021)
湖北省自然科学基金(2012FFB07404)
武汉大学软件工程国家重点实验室开放基金(SKLSE20120934)
中央高校基本科研业务费专项资金项目(CZY12007
CTZ12023)
关键词
局部自适应
女书
字符图像
图像分割
local adaptive
NvShu
character image
image segmentation
