| 基于显著点的感兴趣区域提取及评价 |
| |
| 引用本文： | 陈再良,邹北骥,黄华杰,沈海澜.基于显著点的感兴趣区域提取及评价[J].华中科技大学学报(自然科学版),2012,40(6):65-68. |
| |
| 作者姓名： | 陈再良 邹北骥 黄华杰 沈海澜 |
| |
| 作者单位： | 1. 中南大学信息科学与工程学院,湖南长沙,410083
2. 华南理工大学计算机科学与工程学院,广东广州,510640 |
| |
| 基金项目： | 湖南省自然科学基金资助项目，国家自然科学基金资助项目，中央高校基本科研业务费专项资金资助项目 |
| |
| 摘 要： | 为降低数据规模，提高图像处理的效率，针对感兴趣区域（ROD提取的问题，提出了一种基于显著点的ROI提取算法．利用视觉注意模型（VAM）得到显著图，聚类显著图二值化后提取到的显著点，同时将原图像二值化，并以聚类点作为种子点填充，提取原图像的二值化图轮廓后与原图像掩膜，得到图像的ROI．实验结果表明：该算法的总体效果良好，运算速度可与显著点的个数成线性关系．同时以眼动实验数据为参照进行分析与评价，该算法ROI漏分率至少降低50％，而误分率无明显增加．
|
| 关 键 词： | 图像分割 显著图 显著点 感兴趣区域 眼动 视觉注意模型 |
Saliency point-based extraction and evaluation of interest regions |
| |
| Chen Zailiang,Zou Beiji,Huang Huajie,Shen Hailan.Saliency point-based extraction and evaluation of interest regions[J].JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE,2012,40(6):65-68. |
| |
| Authors: | Chen Zailiang Zou Beiji Huang Huajie Shen Hailan |
| |
| Institution: | 1(1 School of Information Science and Engineering,Central South University,Changsha 410083,China; 2 School of Computer Science and Engineering,South China University of Technology,Guangzhou 510640,China) |
| |
| Abstract: | For the point of reducing the data scale and improving efficiency of image processing,an algorithm of region of interest(ROI) extraction based on saliency point was proposed.The algorithm obtained saliency map with visual attention map(VAM),clusters and optimized saliency points,after binarizing saliency map,and then masks the image after seed filling and source image after its binarizing.Then the ROI can be gotten.The experimental results show that the algorithm can exactly extract ROI in image,and its computing speed is linear with the number of saliency point.Reference to the data of eye movement,performance evaluation shows that the miss classification rate of ROI is decreased by at least 50% and error classification rate does not exhibit further increase. |
| |
| Keywords: | image segmentation saliency map saliency point region of interest eye movement visual attention model |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
