摘要
针对中国书法作品中存在的印章信息难以准确提取的问题,提出基于Guide-filter的中国书法作品中印章信息的提取算法。算法充分利用Guide-filter具有较好保持图像边缘信息的能力,以及最大限度逼近目标信息的特性,以双阈值法得到的二值图像之间的残差图像作为Guide-filter的输入图像,以原始图像作为Guide-filter的引导图像,最终获得较为准确的印章信息。此外,利用边缘密度的概念,通过计算提取印章区域的边缘密度来对印章提取效果进行客观性评价。实验表明:提出算法能够较好地提取出中国书法作品中的印章信息。
In order to solve the problem that the seals are difficult to extract accurately from the Chinese calligraphy works,an algorithm for extraction of seals based on Guide-filter is proposed. This algorithm makes full use of advantage that Guide-filter can preserve the edges of the images,and can approximate the objects greatly.The residual image between binary images obtained by the double threshold values method is used as the input image of Guide-filter,and the original image as guiding image,finally,accurate seal image can be obtained.Besides,the edge density is used to evaluate the effect of seal extraction objectively. The experiment results show that the proposed algorithm can extract the seals information better.
出处
《传感器与微系统》
CSCD
2016年第12期125-128,共4页
Transducer and Microsystem Technologies
基金
国家自然科学基金青年基金资助项目(61502387)
陕西省教育厅专项科研计划资助项目(15JK1748)
西北大学科学研究基金资助项目(14NW25
14NW27
14NW28)