Preview

METHODS OF DISTANCE MEASUREMENT’S ACCURACY INCREASING BASED ON THE CORRELATION ANALYSIS OF STEREO IMAGES

https://doi.org/10.21122/2220-9506-2018-9-1-48-55

Abstract

To solve the problem of increasing the accuracy of restoring a three-dimensional picture of space using two-dimensional digital images, it is necessary to use new effective techniques and algorithms for processing and correlation analysis of digital images. Actively developed tools that allow you to reduce the time costs for processing stereo images, improve the quality of the depth maps construction and automate their construction. The aim of the work is to investigate the possibilities of using various techniques for processing digital images to improve the measurements accuracy of the rangefinder based on the correlation analysis of the stereo image. The results of studies of the influence of color channel mixing techniques on the distance measurements accuracy for various functions realizing correlation processing of images are presented. Studies on the analysis of the possibility of using integral representation of images to reduce the time cost in constructing a depth map areproposed. The results of studies of the possibility of using images prefiltration before correlation processing when distance measuring by stereo imaging areproposed.

It is obtained that using of uniform mixing of channels leads to minimization of the total number of measurement errors, and using of brightness extraction according to the sRGB standard leads to an increase of errors number for all of the considered correlation processing techniques. Integral representation of the image makes it possible to accelerate the correlation processing, but this method is useful for depth map calculating in images no more than 0.5 megapixels. Using of image filtration before correlation processing can provide, depending on the filter parameters, either an increasing of the correlation function value, which is useful for analyzing noisy images, or compression of the correlation function.

About the Author

V. L. Kozlov
Belarusian State University
Belarus

Address for correspondence: Kozlov V.L. - Belarusian State University, Nezavisimosty Ave., 4, Minsk 220050, Belarus.     e-mail: KozlovVL@bsu.by



References

1. Hartley R., Zisserman A. Multiple view geometry in computer vision. Cambridge, Cambridge University Press, 2004, 672 p.

2. Tuytelaars T., Mikolajczyk K. Local invariant feature detectors: A survey. Foundations and Trends in Computer Graphics and Vision, 2008, vol. 3, no. 3, pp. 177–280. doi: 10.1561/0600000017

3. Geusebroek J.M. [et al.] Color invariance. IEEE Transactions on PAMI, 2001, vol. 23, no. 12, pp. 1338– 1350.

4. Lindeberg T. Scale-space theory in computer vision. Dordrecht, Kluwer Academic Publishers, 1994, 69 p.

5. Goshin E.V. A model for the reconstruction of 3D scenes with epipolar constraints taken into account. The young scientist, 2014, no. 12, pp. 71–73 (in Russian). doi: 10.18287/0134-2452-2015-39-5-770-776

6. Ponomarev S.V. A method for comparing stereoscopic algorithms for restoring a three-dimensional model of a person's face. Scientific and Technical Herald of Information Technologies, Mechanics and Optics, 2013, vol. 88, no. 6, pp. 40–45 (in Russian).

7. Simonyan K., Grishin S., Vatolin D. Confidence measure for block-based motion vector field. Computer Graphics and Vision (GraphiCon'2008), Proc. of 18th International Conference, Moscow, 23–27 June 2008, Moscow, 2008, pp. 110–113.

8. Kotyuzhansky L.A. Calculation of the depth map of the stereo image on a graphical processor in real time. Fundamental research, 2012, no. 6, pp. 444–449 (in Russian).

9. Zabih R., Woodfill J. Non-parametric local transforms for computing visual correspondence. Computer Vision ECCV '94: Proc of Third European Conference on Computer Vision, Stockholm, 2–6 May 1994, Springer, 1994, pp. 150–158.

10. JPEG compression standard. Lectures on the course «Methods of information coding». Available at: http://sernam.ru/ lect_cod.php (access: 26.05. 2014).

11. Fisenko V.T., Fisenko T.Yu. Computer processing and image recognition. St. Petersburg, St. Petersburg State University ITMO, 2008, 192 p. (in Russian).

12. Malykhina M.P., Shichkin D.A. Aspects of practical application of color difference for recognition and allocation of image boundaries. Scientific Journal of KubSU, 2013, vol. 89, no. 5, pp. 623–634 (in Russian).

13. Gonzalez R., Woods R. Digital image processing. Moscow, Technosphere Publ., 2005, 1072 p. (in Russian).


Review

For citations:


Kozlov V.L. METHODS OF DISTANCE MEASUREMENT’S ACCURACY INCREASING BASED ON THE CORRELATION ANALYSIS OF STEREO IMAGES. Devices and Methods of Measurements. 2018;9(1):48-55. (In Russ.) https://doi.org/10.21122/2220-9506-2018-9-1-48-55

Views: 3126


Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.


ISSN 2220-9506 (Print)
ISSN 2414-0473 (Online)