Rationale for the use of machine vision for sorting sugar beets when stored in piles

A. I. Zavrazhnov^1,2, S. M. Koltsov^1,2, A. A. Zavrazhnov¹, A. S. Egorov³, D. A. Nikolyukin³, K. A. Manaenkov^1
1Michurinsk State Agrarian University ul. Internatsional’naya, 101, Michurinsk, Tambovskaya obl., 393760, Russian Federation
²All-Russian Research Institute of Application of Equipment and Oil Products in Agriculture, per. Novo-Rubezhnyi, 28, Tambov, 392022, Russian Federation
³Tambov State Technical University, ul. Sovetskaya, 106/5, Tambov, 392000, Russian Federation

Abstract. The studies were carried out to justify the need to develop a software and hardware complex for sorting sugar beets, considering the technological processes of sugar beet production. The period of sugar beet storage in field piles between harvesting in the fields of cultivation and delivery at the plant's beet station can take from several days to several months. However, a long stay in field piles is accompanied by increased losses of beet mass and sucrose, especially in small root crops. Therefore, when forming piles, it is necessary to sort the raw materials. The use of machine vision is relevant for sorting heavy vehicles that deliver raw materials from the fields of cultivation when determining the place of its unloading for storage or processing. The machine vision module, the main control devices of which are a camera, a laser rangefinder and a source of artificial lighting, is installed on the structural elements of the roof of the rejection room, where heavy vehicles are distributed to unloading points. It photographs the surface of the bulk in the body with a camera, on the basis of the information received, the data is processed and the command “for processing” or “for storage” is sent to the display. In addition, the information is sent to the server, where it is analysed graphically and then output to the operator's workstation. One of the approaches to reduce the operational computing power of the microcontroller and improve the accuracy of recognition of individual roots in the bulk of the body of a heavy vehicle involves the use of the method of applying filters to the image. However, its use for solving the problem of sorting sugar beets does not provide the necessary measurement accuracy. Therefore, neural networks should become the main tool for object recognition, and preliminary image preparation by applying filters is advisable to facilitate the work of the neural network.

Keywords: sugar beet; storage; sugar beet root size; machine vision; pile; sorting.

Author Details: A. I. Zavrazhnov, D. Sc. (Tech.), member of the RAS, prof.; S. M. Koltsov, junior research fellow (e-mail: Этот адрес электронной почты защищён от спам-ботов. У вас должен быть включен JavaScript для просмотра.); A. A. Zavrazhnov, Cand. Sc. (Tech.), assoc. prof.; A. S. Egorov, Cand. Sc. (Tech.), assoc. prof.; D. A. Nikolyukin, student; K. A. Manaenkov, D. Sc. (Tech.), prof.

For citation: Zavrazhnov AI, Koltsov SM, Zavrazhnov AA, et al. [Rationale for the use of machine vision for sorting sugar beets when stored in piles]. Dostizheniya nauki i tekhniki APK. 2022;36(12):59-62. Russian. doi: 10.53859/02352451_2022_36_12_59.