Animated Sequences from Image Based Rendering

Description

Promoting products and brands benefit from computer graphics animations. Computer graphics animations make webpages look live. The creation of computer graphics animations usually means large efforts. Image Based Rendering (IBR) is an alternative methodology where base data is collected directly with one or more cameras. In this thesis is explored the use of light-field representations as a way to do animations based on IBR.

Light-field representations (images) are acquired by plenoptic or light-field cameras, which can be thought off as arrays of pinhole cameras. One array of cameras can be represented compactly by an intrinsic matrix [Dansereau13, Zhang18], like what happens with pinhole cameras. A moving camera is an alternative way to acquire light-field (LF) images [Mildenhall19].

Detailed steps:

- LF based generation of image sequences while varying the viewpoint

- LF acquisition with a moving pinhole camera

- Depth based segmentation

- Combination of LFs

References:

[Dansereau13] Dansereau, Donald G., Oscar Pizarro, and Stefan B. Williams. "Decoding, calibration and rectification for lenselet-based plenoptic cameras." Proceedings of the IEEE conference on computer vision and pattern recognition. 2013.

[Zhang18] Zhang, Qi, et al. "A Generic Multi-Projection-Center Model and Calibration Method for Light Field Cameras." IEEE transactions on pattern analysis and machine intelligence (2018).

[Mildenhall19] Local light field fusion: Practical view synthesis with prescriptive sampling guidelines. Mildenhall B, Srinivasan PP, Ortiz-Cayon R, Kalantari NK, Ramamoorthi R, Ng R, Kar A. ACM Transactions on Graphics (TOG). 2019 Jul 12;38(4):1-4. https://github.com/Fyusion/LLFF

Place for conducting the work-proposal:

ISR / IST

More MSc dissertation proposals on Computer and Robot Vision in:

http://users.isr.tecnico.ulisboa.pt/~jag/msc/msc_2021_2022.html