发明授权
- 专利标题: System for improving foggy luminance image using fog reduction estimation model
- 专利标题(中): 使用雾化估计模型改善雾化图像的系统
-
申请号: US14407092申请日: 2013-11-15
-
公开(公告)号: US09418402B2公开(公告)日: 2016-08-16
- 发明人: Sung Hoon Hong , Jae Won Lee
- 申请人: INDUSTRY FOUNDATION OF CHONNAM NATIONAL UNIVERSITY
- 申请人地址: KR Gwangju
- 专利权人: Industry Foundation of Chonnam National University
- 当前专利权人: Industry Foundation of Chonnam National University
- 当前专利权人地址: KR Gwangju
- 代理机构: Rabin & Berdo, P.C.
- 优先权: KR10-2013-0025544 20130311
- 国际申请: PCT/KR2013/010407 WO 20131115
- 国际公布: WO2014/142417 WO 20140918
- 主分类号: G06K9/40
- IPC分类号: G06K9/40 ; G06T5/00 ; G06K9/46 ; G06T7/00 ; G06K9/00
摘要:
The present invention provides a method for providing an image having visibility which is improved by removing fog or smoke from an image of which the quality is reduced due to the fog or the smoke. To this end, the present invention provides an estimation model used for obtaining, from an input image, an image from which the fog is removed, calculates a transmission rate indicating a ratio by which the fog is included in the original image by using the estimation model, and obtains an image of which the fog is removed using the calculated transmission rate. The method proposed in the present invention does not commonly use a filter and only uses calculation of a pixel unit, so that a back lighting effect is not generated, real time processing is possible due to a small amount of calculation, and a good image is obtained even by performing a process using only a luminance image. Therefore, the method can be applied to all multimedia devices in which real time processing is required, such as a monitoring system, a black box, a digital camera, etc.
公开/授权文献
信息查询