David Morales-Norato

Bucaramanga, Colombia

I am a Master Student in Applied Mathematics working in the High Dimensional Signal Processing (HDSP) research group, jointly supervised by Professors Henry Arguello and Hoover Rueda. My research focuses in optics and algorithm design for privacy preserving aware computer vision methods. Before starting my masters, I received my B.S. in Systems Engineer from Universidad Industrial de Santander (UIS) in 2022 supervised by Andrés Jerez and Henry Arguello developing in phase retrieval algorithms.

selected publications

Imaging Congress

Shift-variant PSF characterization of DOE-based imaging systems via implicit neural representations

David Morales-Norato, Henry Arguello, and Hoover Rueda-Chacón

In Optica Imaging Congress (3D, COSI, DH, FLatOptics, IS, pcAOP) 2023

Abs Bib HTML

@inproceedings{cosi_shiftvariant.svg,
  bibtex_show = {true},
  equal = {true},
  abbr = {Imaging Congress},
  author = {Morales-Norato, David and Arguello, Henry and Rueda-Chac\'{o}n, Hoover},
  booktitle = {Optica Imaging Congress (3D, COSI, DH, FLatOptics, IS, pcAOP)},
  journal = {Optica Imaging Congress (3D, COSI, DH, FLatOptics, IS, pcAOP)},
  keywords = {Computational imaging; Diffractive optical elements; Imaging systems; Optical aberration; Optical imaging; Optical systems},
  pages = {CW3B.5},
  publisher = {Optica Publishing Group},
  title = {Shift-variant PSF characterization of DOE-based imaging systems via implicit neural representations},
  year = {2023},
  url = {https://opg.optica.org/abstract.cfm?URI=COSI-2023-CW3B.5},
  html = {https://opg.optica.org/abstract.cfm?URI=COSI-2023-CW3B.5},
  doi = {10.1364/COSI.2023.CW3B.5},
  preview = {cosi_shiftvariant.svg},
  selected = {true}
}

Imaging Congress

Multi-domain Discrete Spectral Image Fusion via Implicit Neural Representation

Manuel Herrera-Poveda, David Morales-Norato, Hans Garcia, and 3 more authors

In Optica Imaging Congress (3D, COSI, DH, FLatOptics, IS, pcAOP) 2023

Abs Bib HTML

@inproceedings{multidomain_fusion.png,
  bibtex_show = {true},
  equal = {true},
  abbr = {Imaging Congress},
  author = {Herrera-Poveda, Manuel and Morales-Norato, David and Garcia, Hans and Castillo, Sergio and Jerez, Andr\'{e}s and Arguello, Henry},
  booktitle = {Optica Imaging Congress (3D, COSI, DH, FLatOptics, IS, pcAOP)},
  journal = {Optica Imaging Congress (3D, COSI, DH, FLatOptics, IS, pcAOP)},
  keywords = {Hyperspectral imaging; Imaging techniques; Neural networks; Optical systems; Quantum efficiency; Sensors},
  pages = {CTh2A.3},
  publisher = {Optica Publishing Group},
  title = {Multi-domain Discrete Spectral Image Fusion via Implicit Neural Representation},
  year = {2023},
  url = {https://opg.optica.org/abstract.cfm?URI=COSI-2023-CTh2A.3},
  html = {https://opg.optica.org/abstract.cfm?URI=COSI-2023-CTh2A.3},
  doi = {10.1364/COSI.2023.CTh2A.3},
  preview = {multidomain_fusion.png},
  selected = {true}
}

Applied Optics
Hyperspectral camera as a compact payload architecture for remote sensing applications

David Morales-Norato, Sergio Urrea, Hans Garcia, and 9 more authors

Applied Optics 2023

Abs Bib HTML

Monitoring and observation over the surface of the Earth have been a matter of global interest. In this path, recent efforts aim to develop a spatial mission to perform remote sensing applications. Mainly, CubeSat nanosatellites have emerged as a standard for developing low-weight and small-sized instruments. In terms of payloads, state-of-the-art optical systems for CubeSats are expensive and designed to work in general use cases. To overcome these limitations, this paper presents a 1.4 U compact optical system to acquire spectral images from a CubeSat standard satellite at the height of 550 km. To validate the proposed architecture, optical simulations using ray tracing simulation software are presented. Because the performance of computer vision tasks is highly related to data quality, we compared the optical system in terms of the classification performance on a real remote sensing application. The performances of the optical characterization and land cover classification show that the proposed optical system achieves a compact instrument, operating at a spectral range from 450 nm to 900 nm discretized on 35 spectral bands. The optical system has an overall f-number of 3.41 with a ground sampling distance of 52.8 m and a swath of 40 km. Additionally, the design parameters for each optical element are publicly available for validation, repeatability, and reproducibility of the results.
@article{cubesat_misc3.png, bibtex_show = {true}, equal = {true}, abbr = {Applied Optics}, author = {Morales-Norato, David and Urrea, Sergio and Garcia, Hans and Rodriguez-Ferreira, Julian and Martinez, Elizabeth and Arguello, Henry and Silva-Lora, Alberto and Torres, Rafael and Acero, Ignacio F. and Hern\'{a}ndez, Francisco L. and C\'{a}rdenas, Lorena P. and Rinc\'{o}n, Sonia}, journal = {Applied Optics}, keywords = {Imaging systems; Machine vision; Optical elements; Optical systems; Ray tracing; Systems design}, number = {8}, pages = {C88--C98}, publisher = {Optica Publishing Group}, title = {Hyperspectral camera as a compact payload architecture for remote sensing applications}, volume = {62}, month = mar, year = {2023}, url = {https://opg.optica.org/ao/abstract.cfm?URI=ao-62-8-C88}, html = {https://opg.optica.org/ao/abstract.cfm?URI=ao-62-8-C88}, doi = {10.1364/AO.476978}, selected = {true} }
Applied Optics

Learning spectral initialization for phase retrieval via deep neural networks

David Morales-Norato, Andrés Jerez, and Henry Arguello

Applied Optics 2022

Abs HTML

Phase retrieval (PR) arises from the lack of phase information in the measures recorded by optical sensors. Phase masks that modulate the optical field and reduce ambiguities in the PR problem by producing redundancy in coded diffraction patterns (CDPs) have been included in these diffractive optical systems. Several algorithms have been developed to solve the PR problem from CDPs. Also, deep neural networks (DNNs) are used for solving inverse problems in computational imaging by considering physical constraints in propagation models. However, traditional algorithms based on non-convex formulation include an initialization stage that requires a high number of iterations to properly estimate the optical field. This work proposes an end-to-end (E2E) approach for addressing the PR problem, which jointly learns the spectral initialization and network parameters. Mainly, the proposed deep network approach contains an optical layer that simulates the propagation model in diffractive optical systems, an initialization layer that approximates the underlying optical field from CDPs, and a double branch DNN that improves the obtained initial guess by separately recovering phase and amplitude information. Simulation results show that the proposed E2E approach for PR requires fewer snapshots and iterations than the state of the art.

Imaging Congress

Deep Phase Retrieval by a Learnable Filtered Spectral Initialization

David Morales-Norato, Andrés Jerez, and Henry Arguello

In OSA Imaging and Applied Optics Congress 2021 (3D, COSI, DH, ISA, pcAOP) 2021

Abs Bib HTML

@inproceedings{cosi_2021.png,
  bibtex_show = {true},
  equal = {true},
  abbr = {Imaging Congress},
  author = {Morales-Norato, David and Jerez, Andr\'{e}s and Arguello, Henry},
  booktitle = {OSA Imaging and Applied Optics Congress 2021 (3D, COSI, DH, ISA, pcAOP)},
  journal = {OSA Imaging and Applied Optics Congress 2021 (3D, COSI, DH, ISA, pcAOP)},
  keywords = {Fourier transforms; Neural networks; Optical imaging; Optical systems; Phase retrieval; Signal processing},
  pages = {CTh4A.6},
  publisher = {Optica Publishing Group},
  title = {Deep Phase Retrieval by a Learnable Filtered Spectral Initialization},
  year = {2021},
  url = {https://opg.optica.org/abstract.cfm?URI=COSI-2021-CTh4A.6},
  html = {https://opg.optica.org/abstract.cfm?URI=COSI-2021-CTh4A.6},
  doi = {10.1364/COSI.2021.CTh4A.6},
  selected = {true}
}