.png){kind=icon}
.png){kind=icon}
Return
Depth Estimation From Light Field by Accumulating Binary Maps Based on Foreground-Background Separation
.png){kind=icon}
PRE
AI
DOI：10.1109/JSTSP.2017.2747154.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Athree-dimensional (3-D) scene can be separated into two regions: foreground and background, which are regions in front of and behind the focused plane, respectively. From the 4-D light field, this paper proposes a depth estimation method by accumulating binary maps, which are computed by the foreground-background separation with light field reparameterization. In the proposed foreground-background separation, an optical phenomenon is used where bundles of rays from the background are flipped on their conjugate planes. Using the Lambertian assumption and gradient constraint, the foreground and background of a scene can be converted to a binary map by voting the gradient signs in every angular patch. Using light field reparameterization, the disparity map can be obtained by accumulating the binary maps. Finding the extremum index in the existing methods corresponds to finding the zero crossing index in the proposed method. By accumulating the binary maps, the proposed method has an advantage on the computational efficiency in terms of memory usage, and can use various reparameterization strategies suitable for applications. Experimental results with synthetic and real images show that the proposed method can generate a high-quality disparity map.
Keywords:
Depth estimation
foreground-background separation
light fields
light field re-parameterization
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
13.7
Papers:
1.9K
Citations:
1.1W