Context
The process of acquiring interferometric hyperspectral images from airborne or satellite sensors [1,2] comprises several steps, including :
- Rectification of the image sequence to a common epipolar geometry to build an epipolar plane image (EPI) sequence.
- Decomposition of each interferometric frame into panchromatic image and fringe pattern.
- Multi-stereoscopic disparity estimation from the panchromatic rectified EPI sequence
- Interpolation of the interferometric pattern for each pixel.
- Reconstruction of the hyperspectral image from the interferometric patterns.
Objective
This project concerns the third step, which is a special kind of subpixel multi-steroscopic problem: The sequence of images are taken from an altitude of 2000m, they have a ground resolution of 50cm, and the translation between one frame and the next in the sequence corresponds to approximately 1 pixel, so each pixel is seen in approximately 400 different images from different angles, which gives extremely low b/h ratios (ranging from 0,1 to 0,00025).
Also, noise and aliasing levels are extremely low for this instrument, and contrast changes are almost negligible, thus providing a quasi-ideal testing ground for subpixel refinement methods as seen in lecture 5.
The objective of this project is to adapt the subpixel refinement method seen in lecture 5 to this case. Adaptations include the generalization from stereo-pairs to multi-stereo sequences, and possibly using optical-flow methods like those described in [3].
Supervision
Andrés Almansa, Gabriele Facciolo
Bibliographic References
[1] Ferrec, Yann, Jean Taboury, Hervé Sauer, Pierre Chavel, Pierre Fournet, Christophe Coudrain, Joël Deschamps, and Jérôme Primot. 2011. “Experimental Results from an Airborne Static Fourier Transform Imaging Spectrometer.” Applied Optics 50 (30): 5894–5904. doi:10.1364/AO.50.005894. [preprint]
[2] Barbanson, Clara, Andrés Almansa, Yann Ferrec, and Pascal Monasse. 2016. “Relief Computation from Images of a Fourier Transform Spectrometer for Interferogram Correction.” In Light, Energy and the Environment, FM3E.6. Washington, D.C.: OSA. doi:10.1364/FTS.2016.FM3E.6. [preprint]
[3] Rais, M., Thiebaut, C., Delvit, J.-M., & Morel, J.-M. (2014). A tight multiframe registration problem with application to Earth observation satellite design. In 2014 IEEE International Conference on Imaging Systems and Techniques (IST) Proceedings (pp. 6–10). IEEE. [doi:10.1109/IST.2014.6958436] [preprint]
