Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Self-Optimization of Camera Hardware

Kristoffersson Lind, Simon LU and Tykesson, Johannes (2021) EDAM05 20211
Department of Computer Science
Abstract
This thesis aims to investigate the automatic tuning of hardware
parameters in a camera's image processing pipeline.
In order to solve the tuning problem,
it is formulated as a black-box optimization problem
centered around a physical camera unit.
Optimization is performed by comparing the camera's output to a reference image.
Several black-box optimization algorithms were tested:
Bayesian Optimization, Evolutionary Optimization,
Particle Swarm Optimization, Simulated Annealing, DIRECT, and Rowan's Subplex Method.
Results indicate that it is feasible to automatically tune
camera hardware parameters using black-box optimization algorithms.
For 14 parameters, Rowan's Subplex Method performs best with an average error of 6.25.
... (More)
This thesis aims to investigate the automatic tuning of hardware
parameters in a camera's image processing pipeline.
In order to solve the tuning problem,
it is formulated as a black-box optimization problem
centered around a physical camera unit.
Optimization is performed by comparing the camera's output to a reference image.
Several black-box optimization algorithms were tested:
Bayesian Optimization, Evolutionary Optimization,
Particle Swarm Optimization, Simulated Annealing, DIRECT, and Rowan's Subplex Method.
Results indicate that it is feasible to automatically tune
camera hardware parameters using black-box optimization algorithms.
For 14 parameters, Rowan's Subplex Method performs best with an average error of 6.25.
When optimizing a much larger set of 71 parameters,
Simulated Annealing, Evolutionary, and Rowan's Subplex Method perform best with an average error of 9.77, 17.92, and 18.05 respectively. (Less)
Popular Abstract
When new cameras are developed, their image quality has to be tuned by expert engineers,
which normally takes several weeks.
Our thesis shows that tuning can be done automatically in a matter of minutes.
Please use this url to cite or link to this publication:
author
Kristoffersson Lind, Simon LU and Tykesson, Johannes
supervisor
organization
alternative title
Automatisk Optimering för Kamerahårdvara
course
EDAM05 20211
year
type
H2 - Master's Degree (Two Years)
subject
keywords
Optimization, Black-box optimization, Evolutionary, Bayesian, Simulated Annealing, DIRECT, Particle Swarm, Simplex
report number
2021-23
ISSN
1650-2884
language
English
id
9057269
date added to LUP
2021-10-11 09:15:50
date last changed
2021-10-11 09:15:50
@misc{9057269,
  abstract     = {{This thesis aims to investigate the automatic tuning of hardware
parameters in a camera's image processing pipeline.
In order to solve the tuning problem,
it is formulated as a black-box optimization problem
centered around a physical camera unit.
Optimization is performed by comparing the camera's output to a reference image.
Several black-box optimization algorithms were tested:
Bayesian Optimization, Evolutionary Optimization,
Particle Swarm Optimization, Simulated Annealing, DIRECT, and Rowan's Subplex Method.
Results indicate that it is feasible to automatically tune
camera hardware parameters using black-box optimization algorithms.
For 14 parameters, Rowan's Subplex Method performs best with an average error of 6.25.
When optimizing a much larger set of 71 parameters,
Simulated Annealing, Evolutionary, and Rowan's Subplex Method perform best with an average error of 9.77, 17.92, and 18.05 respectively.}},
  author       = {{Kristoffersson Lind, Simon and Tykesson, Johannes}},
  issn         = {{1650-2884}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Self-Optimization of Camera Hardware}},
  year         = {{2021}},
}