ACCURACY COMPARISON OF ALGORITHMS FOR DETERMINATION OF IMAGE CENTER COORDINATES IN OPTOELECTRONIC DEVICES

Accuracy in determination of coordinates for image having simple shapes is considered as one of important and significant parameters in metrological optoelectronic systems such as autocollimators, stellar sensors, Shack-Hartmann sensors, schemes for geometric calibration of digital cameras for aerial and space imagery, various tracking systems. The paper describes a mathematical model for a measuring stand based on a collimator which projects a test-object onto a photodetector of an optoelectronic device. The mathematical model takes into account characteristic noises for photodetectors: a shot noise of the desired signal (photon) and a shot noise of a dark signal, readout and spatial heterogeneity of CCD (charge-coupled device) matrix elements. In order to reduce noise effect it is proposed to apply the Wiener filter for smoothing an image and its unambiguous identification and also enter a threshold according to brightness level. The paper contains a comparison of two algorithms for determination of coordinates in accordance with energy gravity center and contour. Sobel, Pruitt, Roberts, Laplacian Gaussian, Canni detectors have been used for determination of the test-object contour. The essence of the algorithm for determination of coordinates lies in search for an image contour in the form of a circle with its subsequent approximation and determination of the image center. An error calculation has been made while determining coordinates of a gravity center for test-objects of various diameters: 5, 10, 20, 30, 40, 50 pixels of a photodetector and also signalto-noise ratio values: 200, 100, 70, 20, 10. Signal-to-noise ratio has been calculated as a difference between maximum image intensity of the test-object and the background which is divided by mean-square deviation of the background. The accuracy for determination of coordinates has been improved by 0.5-1 order in case when there was an increase in a signal-to-noise ratio. Accuracy improvement due to increase of a diameter in a test-object is typical for large signal-to-noise ratios: 70 or more. The conducted investigations have made it possible to establish that the algorithm for determination of coordinates of the energy gravity center is more accurate in comparison with contour methods and requires less computing power (for the MatLab software package), which is related to discreteness while determining a contour.

1. Starosotnikau N. O. (2014) Highly-Accurate Digital Autocollimator for Measuring Small Angles. Novye Napravleniya Razvitiya Priborostroeniya: Materialy 7-i Mezhdunarodnoi Studencheskoi Nauch.-Tekhn. Konferentsii, 23–25 Aprelya 2014 g. [New Directions in Instrumentation Development: Materials of the 7th International Students’ Scientific and Technical Conference]. Minsk, Belarusian National Technical University, 244 (in Russian).

2. Jinyun Yan, Jie Jiang, Guangjun Zhang (2016) Dynamic Imaging Model and Parameter Optimization for a Star Tracker. Optics Express, 24 (6), 5961–5983. DOI: 10.1364/oe.24.005961.

3. Xiaoming Yin, Xiang Li, Liping Zhao, Zhongping Fang (2009) Adaptive Thresholding and Dynamic Windowing Method for Automatic Centroid Detection of Digital Shack–Hartmann Wavefront Sensor. Applied Optics, 48 (32), 6088–6098. DOI: 10.1364/ao.48.006088.

4. Arkhipov S. A., Gasitch G. V., Zavarzin V. I., Morozov S. A. (2008) Photogrammetric Parameters of Optical and Electronic Apparatus. Vestnik Moskovskogo Gosudarstvennogo Tekhnicheskogo Universiteta imeni N. E. Baumana = Herald of the Bauman Moscow State Technical University. Series Instrument Engineering, (4), 105–115 (in Russian).

5. Starosotnikau N. O., Feodortsau R. V. (2015) Estimation of Accurate Determination for Coordinates of Gravity Energy Center in Collimator Test-Object in Respect of Control Schemes for Optoelectronic Devices with Matrix Photodetectors. Nauka i Tekhnika = Science and Technique, (5), 71–76 (in Russian).

6. Zhengrong Huang, Jiangtao Xi, Yanguang Yu, Qinghua Guo (2015) Accurate Projector Calibration Based on a New Point-to-Point Mapping Relationship Between the Camera and Projector Images. Applied Optics, 54 (3), 347–356. DOI: 10.1364/ao.54.000347.

7. Starosotnikau N. O., Feodortsau R. V. (2016) Method for Decreasing Influence of Background Signal Noise while Determining Energy Gravity Centre Coordinates for Images in Electrooptical Devices. Priborostroenie-2016: Materialy 9-i Mezhdunarodnoi Nauch.-Tekhn. Konferentsii, 23–25 Noyabrya 2016 g. [Instrumentation-2016: Materials of the 9th International Scientific and Technical Conference]. Minsk, Belarusian National Technical University, 133–135 (in Russian).

8. Gonzalez R., Woods R., Eddins S. (2004) Digital Image Processing Using MatLab. New Jersey, Prentice Hall.

9. Groen E., Bos J. E., Nacken P.F.M., de Graaf B. (1996) Determination of Ocular Torsion by Means of Automatic Pattern Recognition. IEEE Transactions on Biomedical Engineering, 43 (5), 471–479. DOI: 10.1109/10.488795.

10. Canny J. (1986) A Computational Approach to Edge Detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-8 (6), 679–698. DOI: 10.1109/tpami.1986.4767851.