MATE: Machine Learning for Adaptive Calibration Template Detection
The problem of camera calibration is two-fold. On the one hand, the parameters are estimated from known correspondences between the captured image and the real world. On the other, these correspondences themselves—typically in the form of chessboard corners—need to be found. Many distinct approaches...
| Main Authors: | , , , |
|---|---|
| Format: | Online |
| Language: | English |
| Published: |
MDPI
2016
|
| Online Access: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5134517/ |
| id |
pubmed-5134517 |
|---|---|
| recordtype |
oai_dc |
| spelling |
pubmed-51345172017-01-03 MATE: Machine Learning for Adaptive Calibration Template Detection Donné, Simon De Vylder, Jonas Goossens, Bart Philips, Wilfried Article The problem of camera calibration is two-fold. On the one hand, the parameters are estimated from known correspondences between the captured image and the real world. On the other, these correspondences themselves—typically in the form of chessboard corners—need to be found. Many distinct approaches for this feature template extraction are available, often of large computational and/or implementational complexity. We exploit the generalized nature of deep learning networks to detect checkerboard corners: our proposed method is a convolutional neural network (CNN) trained on a large set of example chessboard images, which generalizes several existing solutions. The network is trained explicitly against noisy inputs, as well as inputs with large degrees of lens distortion. The trained network that we evaluate is as accurate as existing techniques while offering improved execution time and increased adaptability to specific situations with little effort. The proposed method is not only robust against the types of degradation present in the training set (lens distortions, and large amounts of sensor noise), but also to perspective deformations, e.g., resulting from multi-camera set-ups. MDPI 2016-11-04 /pmc/articles/PMC5134517/ /pubmed/27827920 http://dx.doi.org/10.3390/s16111858 Text en © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/). |
| repository_type |
Open Access Journal |
| institution_category |
Foreign Institution |
| institution |
US National Center for Biotechnology Information |
| building |
NCBI PubMed |
| collection |
Online Access |
| language |
English |
| format |
Online |
| author |
Donné, Simon De Vylder, Jonas Goossens, Bart Philips, Wilfried |
| spellingShingle |
Donné, Simon De Vylder, Jonas Goossens, Bart Philips, Wilfried MATE: Machine Learning for Adaptive Calibration Template Detection |
| author_facet |
Donné, Simon De Vylder, Jonas Goossens, Bart Philips, Wilfried |
| author_sort |
Donné, Simon |
| title |
MATE: Machine Learning for Adaptive Calibration Template Detection |
| title_short |
MATE: Machine Learning for Adaptive Calibration Template Detection |
| title_full |
MATE: Machine Learning for Adaptive Calibration Template Detection |
| title_fullStr |
MATE: Machine Learning for Adaptive Calibration Template Detection |
| title_full_unstemmed |
MATE: Machine Learning for Adaptive Calibration Template Detection |
| title_sort |
mate: machine learning for adaptive calibration template detection |
| description |
The problem of camera calibration is two-fold. On the one hand, the parameters are estimated from known correspondences between the captured image and the real world. On the other, these correspondences themselves—typically in the form of chessboard corners—need to be found. Many distinct approaches for this feature template extraction are available, often of large computational and/or implementational complexity. We exploit the generalized nature of deep learning networks to detect checkerboard corners: our proposed method is a convolutional neural network (CNN) trained on a large set of example chessboard images, which generalizes several existing solutions. The network is trained explicitly against noisy inputs, as well as inputs with large degrees of lens distortion. The trained network that we evaluate is as accurate as existing techniques while offering improved execution time and increased adaptability to specific situations with little effort. The proposed method is not only robust against the types of degradation present in the training set (lens distortions, and large amounts of sensor noise), but also to perspective deformations, e.g., resulting from multi-camera set-ups. |
| publisher |
MDPI |
| publishDate |
2016 |
| url |
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5134517/ |
| _version_ |
1613753916783067136 |