钟卿, 余鹏飞, 李海燕, 陈瑞新. 傣文贝叶经的图像增强与二值化方法研究[J]. 云南大学学报(自然科学版), 2017, 39(5): 747-752. doi: 10.7540/j.ynu.20160203
引用本文: 钟卿, 余鹏飞, 李海燕, 陈瑞新. 傣文贝叶经的图像增强与二值化方法研究[J]. 云南大学学报(自然科学版), 2017, 39(5): 747-752. doi: 10.7540/j.ynu.20160203
ZHONG Qing, YU Peng-fei, LI Hai-yan, CHEN Rui-xin. Image enhancement and binarization method research for the Dai palm leaf manuscripts[J]. Journal of Yunnan University: Natural Sciences Edition, 2017, 39(5): 747-752. DOI: 10.7540/j.ynu.20160203
Citation: ZHONG Qing, YU Peng-fei, LI Hai-yan, CHEN Rui-xin. Image enhancement and binarization method research for the Dai palm leaf manuscripts[J]. Journal of Yunnan University: Natural Sciences Edition, 2017, 39(5): 747-752. DOI: 10.7540/j.ynu.20160203

傣文贝叶经的图像增强与二值化方法研究

Image enhancement and binarization method research for the Dai palm leaf manuscripts

  • 摘要: 为了解决贝叶经图像背景不均匀及字符与背景颜色对比度较低给其二值化带来的困难,提出一种图像增强与Sauvola算法相结合的方法.首先使用同态滤波处理贝叶经图像背景不均匀的现象,然后使用灰度线性变换对其进行对比度增强,最后再使用Sauvola算法对贝叶经图像进行二值化.实验结果表明,同态滤波、灰度线性变换与Sauvola算法相结合的方法有效地解决了贝叶经图像背景不均匀和低对比度的问题,得到了清晰的二值化结果,适合于贝叶经图像的图像增强与二值化.

     

    Abstract: In order to solve the problem that the background of Dai palm leaf manuscripts was not uniform and the contrast between character and the background was low,a method combining image enhancement and Sauvola algorithm was proposed.Firstly,the homomorphic filtering was applied to remove the inhomogeneity in the background of Dai palm leaf manuscripts.Subsequently,the linear gray transform was performed to enhance the contrast of the Dai palm leaf manuscripts' image.Finally,the Sauvola algorithm was applied to binarize the Dai palm leaf manuscripts' image.Experiment results showed that the proposed method could solve the uneven background and low contrast of Dai palm leaf manuscripts' image effectively and obtained a clear binary image.Furthermore,the proposed method,combined Sauvola algorithm with the homomorphic filtering and linear gray transform,was more suitable for Dai palm leaf manuscript image binarization compared with the traditional methods.

     

/

返回文章
返回