Compressive spectral imaging using diffractive lenses and multi-spectral sensors with learned reconstruction and joint optimization

Download

Tez_utku_final.pdf

Date

2022-2

Author

Gündoğan, Utku

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

340
views

418
downloads

Compressive spectral imaging aims to reconstruct the entire three-dimensional spectral cube from a few measurements, ideally with a snapshot capability. Recently various spectral imaging modalities have been developed by exploiting diffractive lenses. Another line of development in this area is enabled by spectral filter arrays which resulted in multi-spectral sensors. In this thesis, we first review an existing compressive spectral imaging modality with diffractive lenses and analyze its performance using some quantitative metrics. We then study the joint reconstruction and system optimization for this imaging modality using a model-based learned approach with an unrolled deep neural network (DNN). Secondly, we improve on this modality by developing a new spectral imaging system that also exploits a multi-spectral sensor. To reconstruct the spectral cube from its compressive measurements, we design a learned reconstruction method using a similar unrolled network. A fast sparse reconstruction algorithm is also developed and compared with this learned reconstruction method. The performance of the developed imaging technique is illustrated for the visible regime using different design configurations, number of measurements, and signal-to-noise ratios. The results demonstrate that significant performance improvement can be achieved over the existing compressive spectral imaging modality with diffractive lenses, while also enabling snapshot capability with a simpler design.

Subject Keywords

Multi-spectral sensor, Diffractive lens, Spectral imaging, Compressive sensing, Learned reconstruction, Joint optimization, Inverse problems, Image reconstruction

URI

https://hdl.handle.net/11511/96729

Collections

Graduate School of Natural and Applied Sciences, Thesis

Suggestions

OpenMETU
Core

DEEP LEARNING-BASED UNROLLED RECONSTRUCTION METHODS FOR COMPUTATIONAL IMAGING Bezek, Can Deniz; Öktem, Sevinç Figen; Department of Electrical and Electronics Engineering (2021-9-08) Computational imaging is the process of forming images from indirect measurements using computation. In this thesis, we develop deep learning-based unrolled reconstruction methods for various computational imaging modalities. Firstly, we develop two deep learning-based reconstruction methods for diffractive multi-spectral imaging. The first approach is based on plug-and-play regularization with deep denoisers whereas the second one is an end-to-end learned reconstruction based on unrolling. Secondly, we con...
Compressive spectral imaging with diffractive lenses Kar, Oguzhan Fatih; Öktem, Sevinç Figen (The Optical Society, 2019-09-15) Compressive spectral imaging enables the reconstruction of an entire 3D spectral cube from a few multiplexed images. Here we develop a novel compressive spectral imaging technique using diffractive lenses. Our technique uses a coded aperture to spatially modulate the optical field from the scene and a diffractive lens such as a photon sieve for both dispersion and focusing. Measurement diversity is achieved by changing the focusing behavior of the diffractive lens. The 3D spectral cube is then reconstructed...
Anlık Spektral Görüntüleme için Tasarım Eniyileme Ayazgök, Suleyman; Öktem, Sevinç Figen (2019-08-22) Snapshot spectral imaging enables to reconstructspectral images from a multiplexed single-shot measurement.Since an inversion is required to form the spectral images com-putationally, quantitative characterization of their performanceis essential to optimize the design. In this paper, we analyze theoptimal design of a snapshot spectral imaging technique. Thissnapshot multi-spectral imaging technique uses a diffractive lenscalled generalized photon sieve, and vari...
Compressive Photon-Sieve Spectral Imaging Kar, Oguzhan Fatih; Kamaci, Ulas; Akyon, Fatih; Öktem, Sevinç Figen (2018-06-25) We develop a new compressive spectral imaging modality that utilizes a coded aperture and a photon-sieve for dispersion. The 3D spectral data cube is successfully reconstructed with as little as two shots using sparse recovery
Image fusion for improving spatial resolution of multispectral satellite images Ünlüsoy, Deniz; Süzen, Mehmet Lütfi; Department of Geological Engineering (2013) In this study, four different image fusion techniques have been applied to high spectral and low spatial resolution satellite images with high spatial and low spectral resolution images to obtain fused images with increased spatial resolution, while preserving spectral information as much as possible. These techniques are intensity-hue-saturation (IHS) transform, principle component analysis (PCA), Brovey transform (BT), and Wavelet transform (WT) image fusion. Images used in the study belong to Çankırı reg...

Citation Formats

U. Gündoğan, “Compressive spectral imaging using diffractive lenses and multi-spectral sensors with learned reconstruction and joint optimization,” M.S. - Master of Science, Middle East Technical University, 2022.