基于视觉显著性的人脸文字区域提取  

Regional Extraction of Human Faces and Text Based on Visual Significant

作　　者：何贵阳[1] 

机构地区：[1]重庆城市管理职业学院文化产业管理学院,重庆401331

出　　处：《世界科技研究与发展》2014年第3期268-271,276,共5页World Sci-Tech R&D

摘　　要：基于Matlab图像文字区域检测新方法对图文进行分离和显著性操作,并采用显著性度量对传统人脸文字区域检测的结果做进一步的显著性融合,实现了基于视觉显著性的人脸文字区域提取。实验结果表明,这种基于视觉显著性的人脸文字区域提取解决方案具有比较高的查全率和精确度。Based on the Matlab image text region detection method,a new method combining graphic separation and signifi-cant operation was proposed.Further significant fusion about traditional face text area detection results was realized based on visual significant face text region extraction demonstration system.The experimental results show that the proposed solution has higher recall ratio and precision.

关 键 词：OPENCV 显著性提取 图文分离 C++ 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]