基于图像信息人眼可感知度的图像质量表征方法
文献类型：专利
| 作者 | 朱枫 ; 郝颖明; 吴清潇; 嵇冠群; 吴锦
|
| 发表日期 | 2013-07-03 |
| 专利国别 | 中国 |
| 专利号 | CN103186898B |
| 专利类型 | 发明授权 |
| 产权排序 | 1 |
| 权利人 | 中国科学院沈阳自动化研究所 |
| 其他题名 | Image quality characterization method based on image information human eye perceptible degree |
| 中文摘要 | 本发明公开了基于图像信息人眼可感知度的图像质量表征方法。该方法从人眼视觉特性出发，把图像看成是将实际景物信息传递给人眼的媒介，定义图像质量为人眼可感知的图像信息占图像总信息的百分比。在这一框架下，面向人眼探测识别，提出了3种图像总信息和对应的3种人眼可感知的图像信息的定量表示方法。本发明较一般图像质量表征方法的优点为：将图像质量与人眼能够从图像中接收到的信息关联起来，提出了更符合人眼视觉特性的无参考图像质量表征基本框架；首次提出面向人眼探测识别的图像质量表征方法，可以作为面向探测识别的图像优化的优化准则。 |
| 是否PCT专利 | 否 |
| 英文摘要 | The invention discloses an image quality characterization method based on image information human eye perceptible degree. According to human eye vision characteristics, images are regarded as media which transmit actual scenery information to human eyes, and image quality is determined by the percentage by which image information which can be perceived by the human eyes accounts for image total information. Under the framework, a quantitative representing method of three types of image total information and three corresponding types of image information which can be perceived by the human eyes is provided. Compared with a general image quality characterization method, the image quality characterization method has the advantages that the image quality and the information which can be perceived by the human eyes from the images are related, and the non-reference image quality characterization basic framework which is more in accord with the human eye vision characteristics is put forward the image quality characterization method directing at human eye detection and recognition is put forward for the first time and can be used as an optimization criterion for image optimization directing at detection and recognition. |
| 公开日期 | 2016-01-06 |
| 申请日期 | 2011-12-29 |
| 语种 | 中文 |
| 专利申请号 | CN201110454462.7 |
| 专利代理 | 沈阳科苑专利商标代理有限公司 21002 |
| 源URL | [http://ir.sia.cn/handle/173321/17797]  |
| 专题 | 沈阳自动化研究所_光电信息技术研究室 |
| 推荐引用方式 GB/T 7714 | 朱枫,郝颖明,吴清潇,等. 基于图像信息人眼可感知度的图像质量表征方法. CN103186898B. 2013-07-03. |
入库方式： OAI收割
来源：沈阳自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。