A Detail Enhancement and Dynamic Range Adjustment Algorithm for High Dynamic Range Images
文献类型：会议论文
| 作者 | Xu Bo; Wang Huachuang; Liang Mingtao; Yu Cong; Hu Jinlong; Cheng Hua |
| 出版日期 | 2014 |
| 会议名称 | Proceedings of SPIE: 7TH INTERNATIONAL SYMPOSIUM ON ADVANCED OPTICAL MANUFACTURING AND TESTING TECHNOLOGIES: ADVANCED OPTICAL MANUFACTURING TECHNOLOGIES |
| 会议日期 | 2014 |
| 卷号 | 9281 |
| 页码 | 92811C |
| 通讯作者 | Xu, B (reprint author), Chinese Acad Sci, Inst Opt & Elect, Chengdu 610209, Sichuan, Peoples R China. |
| 中文摘要 | Although high dynamic range (HDR) images contain large amounts of information, they have weak texture and low contrast. What's more, these images are difficult to be reproduced on low dynamic range displaying mediums. If much more information is to be acquired when these images are displayed on PCs, some specific transforms, such as compressing the dynamic range, enhancing the portions of little difference in original contrast and highlighting the texture details on the premise of keeping the parts of large contrast, are needed. To this ends, a multi-scale guided filter enhancement algorithm which derives from the single-scale guided filter based on the analysis of non-physical model is proposed in this paper. Firstly, this algorithm decomposes the original HDR images into base image and detail images of different scales, and then it adaptively selects a transform function which acts on the enhanced detail images and original images. By comparing the treatment effects of HDR images and low dynamic range (LDR) images of different scene features, it proves that this algorithm, on the basis of maintaining the hierarchy and texture details of images, not only improves the contrast and enhances the details of images, but also adjusts the dynamic range well. Thus, it is much suitable for human observation or analytical processing of machines. |
| 英文摘要 | Although high dynamic range (HDR) images contain large amounts of information, they have weak texture and low contrast. What's more, these images are difficult to be reproduced on low dynamic range displaying mediums. If much more information is to be acquired when these images are displayed on PCs, some specific transforms, such as compressing the dynamic range, enhancing the portions of little difference in original contrast and highlighting the texture details on the premise of keeping the parts of large contrast, are needed. To this ends, a multi-scale guided filter enhancement algorithm which derives from the single-scale guided filter based on the analysis of non-physical model is proposed in this paper. Firstly, this algorithm decomposes the original HDR images into base image and detail images of different scales, and then it adaptively selects a transform function which acts on the enhanced detail images and original images. By comparing the treatment effects of HDR images and low dynamic range (LDR) images of different scene features, it proves that this algorithm, on the basis of maintaining the hierarchy and texture details of images, not only improves the contrast and enhances the details of images, but also adjusts the dynamic range well. Thus, it is much suitable for human observation or analytical processing of machines. |
| 学科主题 | Guided filter; detail enhancement; dynamic range adjustment; high dynamic range image processing |
| 语种 | 英语 |
| ISSN号 | 0277-786X |
| 源URL | [http://ir.ioe.ac.cn/handle/181551/7702]  |
| 专题 | 光电技术研究所_光电探测与信号处理研究室（五室） |
| 作者单位 | 1.[Xu Bo 2.Wang Huachuang 3.Yu Cong 4.Hu Jinlong 5.Cheng Hua] Chinese Acad Sci, Inst Opt & Elect, Chengdu 610209, Sichuan, Peoples R China |
| 推荐引用方式 GB/T 7714 | Xu Bo,Wang Huachuang,Liang Mingtao,et al. A Detail Enhancement and Dynamic Range Adjustment Algorithm for High Dynamic Range Images[C]. 见:Proceedings of SPIE: 7TH INTERNATIONAL SYMPOSIUM ON ADVANCED OPTICAL MANUFACTURING AND TESTING TECHNOLOGIES: ADVANCED OPTICAL MANUFACTURING TECHNOLOGIES. 2014. |
入库方式： OAI收割
来源：光电技术研究所
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。