A. I. Tur – Post-graduate Student, Assistant, Department of Automatics and Telemechanics, Perm National Research Polytechnic University

E-mail: tur.aleksandr93@mail.ru

Yu. N. Lipin – Ph.D. (Eng.), Associate Professor, Department of Automatics and Telemechanics, Perm National Research Polytechnic University

E-mail: y_lipinin@mail.ru

Reverse vending machine is an automated way of collecting, sorting and processing used containers (PET bottles and aluminum cans). In the case of visual identification of containers, you should pay attention not only to the shape and color, but also to the packaging elements. The most informative of these elements is the barcode. Recognizing it, we will receive information about the country of production, the name of the organization, the product and even the technical characteristics of the container itself.

At the moment there are already many solutions that allow you to read printable characters (including those based on the TensorFlow used in our project). The neural network can both recognize barcode modules and the numbers under them. Module recognition is performed after reading the extreme lines. They are single segments. With the help of them you can correctly position the barcode in space. The next step is the alternate reading of the remaining lines of unit length. A set of such lines encodes each digit. Digit recognition is performed by the usual printable character recognition algorithm.

However, bottles and cans often give up in a damaged form. Therefore, one should think about creating a system capable of restoring a barcode from a preserved fragment. Barcode redundancy can be used for this (information on successfully read numbers and modules is combined). In the case where both the modules and the corresponding figures are damaged, you can use a mathematical algorithm. Its essence lies in the reverse calculation of the checksum on known figures. In the case of one damaged digit, we are guaranteed to restore the barcode. In other cases, we get a set of possible barcode options. Then with a certain probability it will be possible to guess the barcode.

1. Kokoulin A.N., Tur A.I., Yuzhakov A.A. Convolutional neural networks application in plastic waste recognition and sorting. Proceedings of the 2018 IEEE Conference of Russian Young Researchers in Electrical and Electronic Engineering. Saint-Petersburg: Saint-Petersburg Electrotechnical University «LETI». 2018. P. 1118–1122.
2. Kokoulin A.N., Tur A.I., Dadenkov S.A. Opticheskaya sistema sortirovki pustykh kontejnerov. Nejrokomp'yutery: razrabotka, primenenie. 2018. № 7. S. 3–7. DOI: 10.18127/j19998554-201807-01.
3. Kokoulin A.N., Tur A.I., Knyazev A.I., Yuzhakov A.A. Voprosy razrabotki i nastrojki opticheskoj podsistemy raspoznavaniya i sortirovki pustykh kontejnerov v sisteme razdel'nogo priema bytovykh otkhodov. Informatsionno-izmeritel'nye i upravlyayushchie sistemy. 2018. № 9. T. 16. S. 45–50. DOI: 10.18127/j20700814-201809-07.
4. Blau M. Shtrikh-kody. Kak oni ustroeny? Chast' 1. Ezhednevnyj poznavatel'nyj zhurnal «ShkolaZhizni.ru» [Elektronnyj resurs]. URL: https://shkolazhizni.ru/computers/articles/92623/ (data obrashcheniya: 10.01.19).
5. Blau M. Shtrikh-kody. Kak oni ustroeny? Chast' 2. Ezhednevnyj poznavatel'nyj zhurnal «ShkolaZhizni.ru» [Elektronnyj resurs]. URL: https://shkolazhizni.ru/computers/articles/92624/ (data obrashcheniya: 10.01.19).
6. ZBar bar code reader [Elektronnyj resurs]. URL: http://zbar.sourceforge.net/ (data obrashcheniya: 20.11.18).
7. Ivan'ko M.A., Klepikova A.V. Sistemy iskusstvennogo zreniya. Vestnik MGUP im. Ivana Fedorova. 2015. № 5. S. 50–52.
8. Skaruk G.A. Ierarkhicheskie klassifikatsii v avtomatizirovannom poiske. Trudy GPNTB SO RAN. 2015. № 8. S. 267–274.
9. Favorskaya M.N., Tupitsyn I.V. Ierarkhicheskij metod poiska sootvetstvuyushchikh tochek na stereoizobrazheniyakh. Vestnik Sibirskogo gosudarstvennogo aerokosmicheskogo universiteta im. akademika M.F. Reshetneva. 2012. № 1. S. 62–67.
10. Kokoulin A.N. Ispol'zovanie nejronnykh setej dlya obnaruzheniya i raspoznavaniya pylevykh chastits na mikrofotografiyakh. Nejrokomp'yutery: razrabotka, primenenie. 2015. № 10. S. 10–15.