摘要
针对复杂图片中文字对视觉的干扰,通过图像分解对其结构分量与纹理分量进行分离,结构分量中的文字干扰采用TV模型对其进行消除,纹理分量中的文字干扰运用Laplacian算子对所在区域进行定位,再用Criminisi算法对其有效消除.仿真结果表明,该方法对复杂图片中文字的干扰消除高效实用.
For the visual interference of text in complex images,the structural components and texture components are separated by image decomposition.The text interference in structural components is eliminated by TV model.The Laplacian operator is used to locate the region where the text interference is,and then we effectively eliminate it with the Criminisi algorithm.The simulation results show that the method is effective for eliminating the interference of texts in complex images.
作者
董剑龙
王浩全
DONG Jianlong;WANG Haoquan(Academic Affairs Office,Taiyuan Institute of Technology,Taiyuan Shanxi030008,China;Key Laboratory of Instrumentation Science&Dynamic Measurement,Ministry of Education,North University of China,Taiyuan Shanxi030051,China)
出处
《新疆大学学报(自然科学版)》
CAS
2019年第3期330-334,共5页
Journal of Xinjiang University(Natural Science Edition)
基金
山西省回国留学人员科研资助项目(2017-089)
山西省自然科学基金项目(201801D121160)