针对影视作品中的大量离焦图像,提出了一种离焦图像的深度图提取方法.将离焦图像的聚焦前景和离焦背景进行分离.对离焦背景提出了深度图模型匹配的方法,构建深度图模型并结合人眼视觉对场景深度的敏锐判断,将背景与对应的深度图模型进行匹配,实现背景深度图的构建；提出了基于颜色分割的深度图再处理,来进一步提高场景深度图的精度.对前景采用单深度赋值,并结合背景深度图融合生成最终深度图.实验表明采用该方法提取的深度图在深度跳跃和深度平滑区域都得到了好的效果.
参考文献
| [1] | Wong K T,Ernst F.Master thesis “Single Image Depth-from-Defocus”[D].Eindhoven,The Netherland:Delft University of Technology & Philips Natlab Research,2004. |
| [2] | Valencia S A,Rodriguez-Dagnino R M.Synthesizing stereo 3D view from focus cues in monoscopic 2D images[J].Proc.SPIE,2003,5006:377-388. |
| [3] | Guo G,Zhang N,Huo L,et al.2D to 3D Conversion based on edge defocus and segmentation[C]//IEEE International Conference on Acoustics,Speech and Signal Processing,Las Vegas,America:IEEE,2008:2181-2184. |
| [4] | Chen Y C,Wu Y C,Liu C H,et al.Depth map generation based on depth from focus[C]//IEEE International Conference on Systems and Applications,Kuala Lumpur,Malaysia:IEEE,2010:59-63. |
| [5] | Li H (l),Ngan K N.Unsupervized video segmentation with low depth of field[J].IEEE Transactions on Circuits and Systems for Video Technology,2007,17(12):1742-1751. |
| [6] | Comaniciu D,Meer P.Mean shift-a robust approach toward feature space analysis[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(5):603-619. |
| [7] | Battiato S,Curti S,La Cascia M,et al.Depth-map generation by image classification[C]//SPIE IS& T/SPIE's 16th Annual Symposium on Electronic Imaging,San Jose,America:SPIE,2004. |
| [8] | Nedovic V,Smeulders A W M,Redert A,et al.Depth Estimation via Stage Classification[C]//3DTV-Conference,Istanbul,Turkey:IEEE,2008:77-80. |
| [9] | Zhang R,Tsai P,Cryer J E,et al.Shape from shading:A survey[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1999,21(8):690-706. |
| [10] | Kang G W,Gan C Q,Ren W W.Shape from shading based on finite-element[C]//Proceedings of International Conference on Machine Learning and Cybernetics,Guangzhou,China:IEEE,2005,8:5165-5169. |
上一张
下一张
上一张
下一张
计量
- 下载量()
- 访问量()
文章评分
- 您的评分：
-
10%
-
20%
-
30%
-
40%
-
50%