研究目的
Investigating the inverse af?ne Radon transform for reconstructing the light ?eld from the focal stack.
研究成果
The high-precision light ?eld data can be reconstructed via the approximation of the inverse af?ne Radon transform. The light ?eld image of a single perspective has the large depth of ?eld. The high precision scene depth with pixel-by-pixel can be reconstructed via the light ?eld data.
研究不足
The partial derivative operator is sensitive to noise, and the denominator of the integral function has singularities, revealing the ill-posedness of the inverse transform. Another reason for ill-posedness is the incompleteness of the focal stack data.
1：Experimental Design and Method Selection:
The study employs the af?ne Radon transform to model the generation of the focal stack from the light ?eld and derives its inverse for light field reconstruction.
2：Sample Selection and Data Sources:
The experiments use focal stack data captured by a Point Grey camera with a Myutron lens.
3：List of Experimental Equipment and Materials:
Point Grey camera (model: GS3-U3-60S6M-C) and Myutron lens (model: HF5018V).
4：Experimental Procedures and Operational Workflow:
The focal stack data is used to reconstruct the light field and further reconstruct the depth of the scene for 3D scene reconstruction.
5：Data Analysis Methods:
The numerical solution of light field reconstruction is realized by approximating the inverse of af?ne Radon transform.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
