计算机技术与发展Issue(7)：234-237,4.DOI:10.3969/j.issn.1673-629X.2015.07.053
基于改进视觉注意模型的显著区域区检测
Region of Interest Detection Based on Improved Visual Attention Model
摘要
Abstract
In this paper,propose a new visual attention model improvement algorithm based on the visual attention model proposed by ex-perts. First,the improved Itti model is used to compute the visual saliency of image,from low-level features including color,brightness and direction,then construct a wavelet Pyramid,using the central peripheral operator for difference operation to attain the concern graph of each feature of color,brightness and direction,and then through the normalized operator,merging into the saliency map of Itti model. Next,drawing on the experience of algorithm proposed by MA and Zhang etc,calculate the saliency map from a global perspective. Final-ly the saliency map above are normalized fusion.关键词
视觉注意模型/Itti模型/显著图/多尺度变换Key words
visual attention model/Itti model/saliency/multi-scale transform分类
信息技术与安全科学引用本文复制引用
宋侃..基于改进视觉注意模型的显著区域区检测[J].计算机技术与发展,2015,(7):234-237,4.基金项目
安徽省科技攻关计划科技强警专项资金资助项目(1301b042020) （1301b042020）
安徽大学青年骨干教师培养对象经费资助项目 （）
高等学校博士学科点专项科研基金联合资助课题(20133401110009) （20133401110009）
