Mathematical model and a parallel algorithm for processing images containing symbolic information about products

Аuthors

Khomyakov O. O.^1*, Panishchev V. S.^1**, Titov V. S.^1***, Vatutin E. I.^2****

1. South-Western State University, 94, 50-let Oktyabrya str., Kursk, 305040, Russia
2. ,

*e-mail: homyakov46rus@yandex.ru
**e-mail: gskunk@yandex.ru
***e-mail: titov-kstu@rambler.ru
****e-mail: evatutin@rambler.ru

Abstract

The purpose of the study consists in developing a mathematical model and algorithm for parallel processing of symbolic information on the product labels to increase the character recognition accuracy in the presence of noise and distortion.
The authors employed image processing methods for image regions highlighting, distortions elimination and noise compensation. A neural network approach was applied for the text search and recognition. Methods of parallel computing systems designing were employed as well, and computer modeling methods were applied for testing the developed solution. The authors analyzed methods employed for video stream frames processing to highlight certain objects, and methods used for digital image processing, particularly for the text information extraction. An algorithm for the parts of text information classification based on reference features has been developed. These methods and algorithms have been tested under various conditions. The character classifier implementation was compared with an alternative version. Examples and results of the developed system are demonstrably presented. The proposed solution can be successfully integrated into production process automation systems. It is suitable for a wide range of applications, including product quality control, inventory and warehouse operations automation. The study demonstrates that compared to the analogues, the developed model and algorithm are of higher accuracy and throughput at video streams processing and symbolic information recognition. Compared to the analog, the accuracy increase was from 17% to 30%, depending on the noise level in the image, and the bandwidth was increased by 50%.

Keywords:

character information recognition, OCR, image preprocessing, video stream processing, text information extraction, classification, image recognition, labeling

References

1. Kopylov I., Kazakov A., Malygin L. Vestnik Cherepovetskogo gosudarstvennogo universiteta, 2016, no. 74, pp. 12-15.
2. Provotorov A.V., Orlov A.A. Sovremennye problemy nauki i obrazovaniya, 2012, no. 6, pp. 98.
3. Pyzh S.V., Ganicheva O.G. Mobil'naya avtomatizirovannaya sistema inventarizatsii metallurgicheskoi produktsii i obespecheniya bezopasnosti v skladskikh pomeshcheniyakh. Nauchno-tekhnicheskii progress v chernoi metallurgii (Mobile automated system for inventory of metallurgical products and security in warehouses. Scientific and technical progress in ferrous metallurgy): sbornik statei. Cherepovets: Cherepovetskii gosudarstvennyi universitet, 2013. pp. 266-271.
4. Panishchev V.S., Trufanov M.I., Dobroserdov O.G., Khomyakov O.O. Izvestiya Yugo-Zapadnogo gosudarstvennogo universiteta, 2021, no. 1, pp. 122-137. DOI: 10.21869/2223-1560-2020-25-1-122-137
5. Chung I., Sainath T., Ramabhadran B., Pichen M. et al. Parallel Deep Neural Network Training for Big Data on Blue GeneQ, International Conference for High Performance Computing, Networking, Storage and Analysis, SC, 2015, pp. 745-753. DOI: 10.1109/SC.2014.66
6. Qiu Q., Wu Q., Bishop M., Pino R., Linderman R. A Parallel Neuromorphic Text Recognition System and Its Implementation on a Heterogeneous High-Performance Computing Cluster, IEEE Transactions on Computers, 2013, vol. 62, pp. 886-899. DOI: 10.1109/TC.2012.50
7. Ol'kina D.S. Trudy MAI, 2023, no. 130. URL: https://trudymai.ru/eng/published.php?ID=174617. DOI: 10.34759/trd-2023-130-18
8. Malygin I.V., Bel'kov S.A., Tarasov A.D., Usvyatsov M.R. Trudy MAI, 2017, no. 96. URL: https://trudymai.ru/eng/published.php?ID=85797
9. Khomonenko A.D., Yakovlev E.L. Naukoemkie tekhnologii v kosmicheskikh issledovaniyakh Zemli, 2018, no. 6, pp. 86-93.
10. Lukezic A., Voj'ir T., Cehovin Zajc L., Matas J., Kristan M. Discriminative correlation filter tracker with channel and spatial reliability // International Journal of Computer Vision, 2018, pp. 6309-6318. DOI: 10.1007/s11263-017-1061-3
11. Zivkovic Z., Taylan Cemgil A., Kröse B. Approximate Bayesian methods for kernel-based object tracking. Computer Vision and Image Understanding, 2009, pp. 743-749.
12. Rublee E., Rabaud V., Konolige K., Bradski G. Orb: an efficient alternative to sift or surf. Computer Vision (ICCV), 2011 IEEE International Conference, 2011, pp. 2564–2571.
13. Alekseev V.V., Lakomov D.V. Trudy MAI. 2018, no. 103. URL: https://trudymai.ru/eng/published.php?ID=100810
14. Zhong X., You Z., Qian M., Zhang J., Hu X. Metal stamping character recognition algorithm based on multi-directional illumination image fusion enhancement technology, EURASIP Journal on Image and Video Processing, 2018, pp. 1-11. DOI: 10.1186/s13640-018-0321-7
15. Gonsales R., Vuds R. Tsifrovaya obrabotka izobrazhenii (Digital image processing), Moscow, Tekhnosfera, 2012, 1104 p.
16. Shapiro L., Stokman Dzh. Komp'yuternoe zrenie (Computer vision), Moscow, BINOM, 2013, 752 p.
17. Ramer U. An iterative procedure for the polygonal approximation of plane curves, Computer Graphics and Image Processing, 1972, pp. 244-256. DOI:10.1016/S0146-664X(72)80017-0
18. Douglas D.H., Peucker T.K. Algorithms for the reduction of the number of points required to represent a digitized line or its caricature. Geovisualization, University of Toronto Press, 1973, pp. 112-122.
19. Zhou X., Yao C., Wen H., Wang Y., Zhou S., He W., Liang J. EAST: An Efficient and Accurate Scene Text Detector, Computer Vision and Pattern Recognition (CVPR), 2017. DOI: 10.1109/CVPR.2017.283
20. Silakov N.V., Tassov K.L. Vestnik RGGU. Seriya: Informatika. Informatsionnaya bezopasnost'. Matematika, 2020, no. 2, pp. 27-45. DOI: 10.28995/2686-679X-2020-2-27-45

Download