DETECTION OF BACTERIA IN FOODSTUFF BY MACHINE LEARNING METHODS

A. . Saenko, V. M. Musalimov, S. . Lerm, G. . Linss


Read the full article 

Abstract

The paper deals with an actual problem of ensuring the control of foodstuff quality by means of machine learning methods. Existing analysis methods require special laboratory environment, significant time and depend on the qualification and some physiological characteristics of an expert while the suggested method gives the possibility to decrease significantly the costs due to automatization. The mobile analysis platform performing this method is based on the fluorescence microscopy. The problem of the object classification as either “bacterium” or “third-party artifact” was solved for the test data with some classification algorithms as support vector machine, random forest, decision tree C4.5, k-nearest neighbors, Bayes method. The analysis showed that the most effective algorithms are support vector machine and random forest. This research is performed on the Mechatronics Department of Saint Petersburg National Research University of Information Technologies, Mechanics and Optics and the Quality Assurance and Industrial Image Processing Department of Ilmenau University of Technology with the support of the program “Mikhail Lomonosov” of the Ministry of Education and Science of Russia and the German Academic Exchange Service.


Keywords: machine learning, bacteria detection

References
1. Bigus J.P. Data Mining with Neural Networks. McGraw-Hill, 1996. P. 220.
2. The handbook of data mining / Ed. N.Ye. Lawrence Erlbaum Associates, 2003. 689 p.
3. Дьяконов А.Г. Анализ данных, обучение по прецедентам, логические игры, системы WEKA, RapidMiner и MatLab (Практикум на ЭВМ кафедры математических методов прогнозирования): Учеб.пособие. М.: Издательский отдел факультета ВМК МГУ им. М.В. Ломоносова, 2010. 278 с 
4. Anding K. Automatisierte Qualitätssicherung von Getreide mit überwachten Lernverfahren in der Bildverarbeitung: Dissertation zur Erlangung der akademischen Grades Doktoringenieur (Dr.-Ing.). Ilmenau, Germany: Technische Universität Ilmenau, 2010. 235 p.
5. Witten I.H., Frank E., Hall M.A. Data Mining: Practical Machine Learning Tools and Techniques. 3rd ed. Morgan Kaufmann, 2011. 629 p.
6. Bramer M. Principles of data mining. 2nd ed. Springer, 2013. 440 p.
7. Латыев С.М., Воронин А.А., Андинг К., Линц Э., Курицын П.А. Оптико-электронные методы и средства идентификации веществ и материалов // Изв. вузов. Приборостроение. 2013. Т. 56. № 10. С. 81–87.
8. ISO 6579:2002. Microbiology of food and animal feeding stuffs – Horizontal method for the detection of Salmonella spp. 08.08.2002. 32 p.
9. Lerm S., Holder S., Gopfert A., Futterer R., Linss G. Concepts of a scanning hardware platform for highresolution image processing with Lab-on-a-chip analysis // Proc. of the 15th International Symposium «MECHATRONIKA». Prague, 2012. P. 1–4.
10. Lerm S. Objektsegmentierung von kompakten Schüttgut für die technische Erkennung: Dissertation zur Erlangung der akademischen Grades Doktoringenieur (Dr.-Ing.). Ilmenau, Germany: Technische Universität Ilmenau, 2012. 235 p.
11. Саенко А.П. Программная система бинаризации и сегментации изображений: Свидетельство о государственной регистрации программы для ЭВМ 2013613134. Заявл. 2013.
12. Shapiro L., Stockman G. Computer Vision. Prentice Hall PTR, 2001. 580 p.
13. Gonzalez R.C., Woods R.E. Digital Image Processing. 2nd ed. Prentice Hall, 2002. 793 p.
14. Erhardt A. Einführung in die Digitale Bildverarbeitung. Vieweg+Teubner Verlag, 2008. 248 р.
15. HALCON Version 11.0.1 – HALCON / HDevelop Reference Manual. MVTec Software GmbH, 2012. 2352 p.
16. Саенко А.П. Оценка эффективности обнаружения бактерий методами обработки цифровых изображений и интеллектуального анализа данных // Сборник научных трудов Одиннадцатой сессии международной научной школы «Фундаментальные и прикладные проблемы надежности и диагностики машин и механизмов». СПб: Институт проблем машиноведения РАН, 2013. С. 318–321.
Copyright 2001-2017 ©
Scientific and Technical Journal
of Information Technologies, Mechanics and Optics.
All rights reserved.

Яндекс.Метрика