Robust camera calibration for the MiroSot and the AndroSot vision systems using artificial neural networks

Awang Hendrianto Pratomo, Mohamad Shanudin Zakaria, Mohammad Faidzul Nasrudin, Anton Satria Prabuwono, Choong Yeun Liong, Izwan Azmi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

The MirosSot and the AndroSot soccer robots have the ability to recognize, and navigate within, their environments without human intervention. An overhead global camera, usually at a fixed position, is used for the robot’s vision. Because of the lens distortion, images obtained from the camera do not accurately represent the robot’s environment. The distortions affect the coordinates. A technique to calibrate the camera is required to transform the skewed coordinates of the objects in the image to the physical coordinates, which define their real-world position. In this study, a method is proposed for camera calibration using an artificial neural network (ANN) in a two-step process. First, ANN was used to select the camera height and the lens focal lengths for high accuracy. Second, ANN was used to map a coordinate transformation from the camera coordinates to the physical coordinates. During the learning process, the weight of each node in the ANN model changed until the best architecture is reached. The experiments thus resulted in an optimum ANN architecture of 2×4×25×2. The accuracy and efficiency of the camera calibration method were obtained by relearning using the ANN whenever changes to the environmental occurred. Relearning was done using the new input data set for each respective environmental change. Based on our experiments, the average transformation error of the calibration method, using many types of camera, camera positions, camera heights, lens sizes, and focal lengths, was 0.18283 cm.

Original languageEnglish
Title of host publicationAdvances in Intelligent Systems and Computing
PublisherSpringer Verlag
Pages571-585
Number of pages15
Volume345
ISBN (Print)9783319168401
DOIs
Publication statusPublished - 2015
Event3rd International Conference on Robot Intelligence Technology and Applications, RiTA 2014 - Beijing, China
Duration: 6 Nov 20148 Nov 2014

Publication series

NameAdvances in Intelligent Systems and Computing
Volume345
ISSN (Print)21945357

Other

Other3rd International Conference on Robot Intelligence Technology and Applications, RiTA 2014
CountryChina
CityBeijing
Period6/11/148/11/14

Fingerprint

Cameras
Calibration
Neural networks
Lenses
Robots
Camera lenses
Network architecture
Computer vision
Experiments

Keywords

  • Camera calibration technique
  • Global vision system
  • Neural network
  • Robot soccer

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Computer Science(all)

Cite this

Pratomo, A. H., Zakaria, M. S., Nasrudin, M. F., Prabuwono, A. S., Liong, C. Y., & Azmi, I. (2015). Robust camera calibration for the MiroSot and the AndroSot vision systems using artificial neural networks. In Advances in Intelligent Systems and Computing (Vol. 345, pp. 571-585). (Advances in Intelligent Systems and Computing; Vol. 345). Springer Verlag. https://doi.org/10.1007/978-3-319-16841-8_51

Robust camera calibration for the MiroSot and the AndroSot vision systems using artificial neural networks. / Pratomo, Awang Hendrianto; Zakaria, Mohamad Shanudin; Nasrudin, Mohammad Faidzul; Prabuwono, Anton Satria; Liong, Choong Yeun; Azmi, Izwan.

Advances in Intelligent Systems and Computing. Vol. 345 Springer Verlag, 2015. p. 571-585 (Advances in Intelligent Systems and Computing; Vol. 345).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Pratomo, AH, Zakaria, MS, Nasrudin, MF, Prabuwono, AS, Liong, CY & Azmi, I 2015, Robust camera calibration for the MiroSot and the AndroSot vision systems using artificial neural networks. in Advances in Intelligent Systems and Computing. vol. 345, Advances in Intelligent Systems and Computing, vol. 345, Springer Verlag, pp. 571-585, 3rd International Conference on Robot Intelligence Technology and Applications, RiTA 2014, Beijing, China, 6/11/14. https://doi.org/10.1007/978-3-319-16841-8_51
Pratomo AH, Zakaria MS, Nasrudin MF, Prabuwono AS, Liong CY, Azmi I. Robust camera calibration for the MiroSot and the AndroSot vision systems using artificial neural networks. In Advances in Intelligent Systems and Computing. Vol. 345. Springer Verlag. 2015. p. 571-585. (Advances in Intelligent Systems and Computing). https://doi.org/10.1007/978-3-319-16841-8_51
Pratomo, Awang Hendrianto ; Zakaria, Mohamad Shanudin ; Nasrudin, Mohammad Faidzul ; Prabuwono, Anton Satria ; Liong, Choong Yeun ; Azmi, Izwan. / Robust camera calibration for the MiroSot and the AndroSot vision systems using artificial neural networks. Advances in Intelligent Systems and Computing. Vol. 345 Springer Verlag, 2015. pp. 571-585 (Advances in Intelligent Systems and Computing).
@inproceedings{2b27c892c18f429aa62101f433f95017,
title = "Robust camera calibration for the MiroSot and the AndroSot vision systems using artificial neural networks",
abstract = "The MirosSot and the AndroSot soccer robots have the ability to recognize, and navigate within, their environments without human intervention. An overhead global camera, usually at a fixed position, is used for the robot’s vision. Because of the lens distortion, images obtained from the camera do not accurately represent the robot’s environment. The distortions affect the coordinates. A technique to calibrate the camera is required to transform the skewed coordinates of the objects in the image to the physical coordinates, which define their real-world position. In this study, a method is proposed for camera calibration using an artificial neural network (ANN) in a two-step process. First, ANN was used to select the camera height and the lens focal lengths for high accuracy. Second, ANN was used to map a coordinate transformation from the camera coordinates to the physical coordinates. During the learning process, the weight of each node in the ANN model changed until the best architecture is reached. The experiments thus resulted in an optimum ANN architecture of 2×4×25×2. The accuracy and efficiency of the camera calibration method were obtained by relearning using the ANN whenever changes to the environmental occurred. Relearning was done using the new input data set for each respective environmental change. Based on our experiments, the average transformation error of the calibration method, using many types of camera, camera positions, camera heights, lens sizes, and focal lengths, was 0.18283 cm.",
keywords = "Camera calibration technique, Global vision system, Neural network, Robot soccer",
author = "Pratomo, {Awang Hendrianto} and Zakaria, {Mohamad Shanudin} and Nasrudin, {Mohammad Faidzul} and Prabuwono, {Anton Satria} and Liong, {Choong Yeun} and Izwan Azmi",
year = "2015",
doi = "10.1007/978-3-319-16841-8_51",
language = "English",
isbn = "9783319168401",
volume = "345",
series = "Advances in Intelligent Systems and Computing",
publisher = "Springer Verlag",
pages = "571--585",
booktitle = "Advances in Intelligent Systems and Computing",

}

TY - GEN

T1 - Robust camera calibration for the MiroSot and the AndroSot vision systems using artificial neural networks

AU - Pratomo, Awang Hendrianto

AU - Zakaria, Mohamad Shanudin

AU - Nasrudin, Mohammad Faidzul

AU - Prabuwono, Anton Satria

AU - Liong, Choong Yeun

AU - Azmi, Izwan

PY - 2015

Y1 - 2015

N2 - The MirosSot and the AndroSot soccer robots have the ability to recognize, and navigate within, their environments without human intervention. An overhead global camera, usually at a fixed position, is used for the robot’s vision. Because of the lens distortion, images obtained from the camera do not accurately represent the robot’s environment. The distortions affect the coordinates. A technique to calibrate the camera is required to transform the skewed coordinates of the objects in the image to the physical coordinates, which define their real-world position. In this study, a method is proposed for camera calibration using an artificial neural network (ANN) in a two-step process. First, ANN was used to select the camera height and the lens focal lengths for high accuracy. Second, ANN was used to map a coordinate transformation from the camera coordinates to the physical coordinates. During the learning process, the weight of each node in the ANN model changed until the best architecture is reached. The experiments thus resulted in an optimum ANN architecture of 2×4×25×2. The accuracy and efficiency of the camera calibration method were obtained by relearning using the ANN whenever changes to the environmental occurred. Relearning was done using the new input data set for each respective environmental change. Based on our experiments, the average transformation error of the calibration method, using many types of camera, camera positions, camera heights, lens sizes, and focal lengths, was 0.18283 cm.

AB - The MirosSot and the AndroSot soccer robots have the ability to recognize, and navigate within, their environments without human intervention. An overhead global camera, usually at a fixed position, is used for the robot’s vision. Because of the lens distortion, images obtained from the camera do not accurately represent the robot’s environment. The distortions affect the coordinates. A technique to calibrate the camera is required to transform the skewed coordinates of the objects in the image to the physical coordinates, which define their real-world position. In this study, a method is proposed for camera calibration using an artificial neural network (ANN) in a two-step process. First, ANN was used to select the camera height and the lens focal lengths for high accuracy. Second, ANN was used to map a coordinate transformation from the camera coordinates to the physical coordinates. During the learning process, the weight of each node in the ANN model changed until the best architecture is reached. The experiments thus resulted in an optimum ANN architecture of 2×4×25×2. The accuracy and efficiency of the camera calibration method were obtained by relearning using the ANN whenever changes to the environmental occurred. Relearning was done using the new input data set for each respective environmental change. Based on our experiments, the average transformation error of the calibration method, using many types of camera, camera positions, camera heights, lens sizes, and focal lengths, was 0.18283 cm.

KW - Camera calibration technique

KW - Global vision system

KW - Neural network

KW - Robot soccer

UR - http://www.scopus.com/inward/record.url?scp=84942626855&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84942626855&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-16841-8_51

DO - 10.1007/978-3-319-16841-8_51

M3 - Conference contribution

AN - SCOPUS:84942626855

SN - 9783319168401

VL - 345

T3 - Advances in Intelligent Systems and Computing

SP - 571

EP - 585

BT - Advances in Intelligent Systems and Computing

PB - Springer Verlag

ER -