CC BY
1Prototype of an optical system for identifying micro- and macrodamage to plant tissues
A.Yu. Izmailov*, A.S. Dorokhov, A.V. Sibirev
1- FSBSI "Federal Scientific Agronomic and Engineering Center VIM", Russia, Moscow, 109428
When storing fruit and vegetable products, one should note the circumstance caused by the natural loss of potato weight depending on the degree of damage to the tubers, which ranges from 11% to 22%, and waste in the form of rot is 7...25%, while for potatoes without mechanical damage, natural loss is 5%, and waste is 2...4% [1,2]. To meet these standards and ensure product quality, the development of non-invasive, high-throughput methods for recognizing and classifying fruits and vegetables, identifying their disease damage during sorting in real time is required. In order to automatically recognize infected and damaged vegetable crops, potatoes and apple fruits during post-harvest processing, together with Sapsan LLC (Voronezh), an optical system for identifying biological objects has been developed using laser and spectral technologies to obtain correlation-spectral patterns. The optical identification system is intended for express diagnostics of the functional state of vegetable crops, potatoes and apple fruits during post-harvest processing [3,4]. The optical identification system provides the required performance indicators for recognizing biological objects based on discriminant functions with multi-class recognition using the minimal risk method and determining the quantitative characteristics of the objects under study. The optical system for identifying micro- and macro-damage to fruits and vegetables is a set of modules for feeding, optical and sorting biological objects (fruits and vegetables). The module for feeding fruits and vegetables consists of a conveyor with a width of at least 300 mm, made of polymer/rubber material and made with a directed relief for uniform distribution of biological objects over the working surface with an adjustable feed speed from 0.5 to 3.0 m/s. The optical module is a system consisting of technical vision devices and a control unit with a system for illuminating biological objects, with the ability to move above the surface of the conveyor-sorting adapter. A technical vision system consists of at least 2 cameras with a resolution of at least 5400 pixels and operating modes in the RGB+RGB or NIR+RGB spectral range. The sorting module consists of a conveyor-sorting adapter with 64 pneumatic actuators installed on the working surface and an air supply pressure in the range of 0.2 - 0.8 MPa with a maximum air flow of at least 1000 l/min. The working surface of the fruit and vegetable feeding module must be made of polymer/rubber material. The novelty of the development lies in obtaining correlation-spectral patterns for recognizing contaminated vegetable crops, potatoes and apple fruits using optical technologies that determine the physiological state and topological parameters of biological objects based on elements of artificial intelligence and machine learning. The uniqueness of the system for identifying infected biological objects lies in the presence of adaptive disease recognition devices depending on the culture using specialized optical systems. The accuracy of recognition of qualified biological objects is more than 95% of those affected by diseases. This ensures the possibility of obtaining potato seeds, vegetable crops and apple fruits in the absence of foci of mechanical damage and disease on their surface after storing them. The economic efficiency of the optical system for identifying micro- and macro-damage during the sorting operation makes it possible to reduce the number of personnel involved in the post-harvest processing stage, while increasing productivity by 1.8 times, as well as the quality of sorting of marketable products as a result of disease recognition during remote optical monitoring and subsequent rejection of substandard products. The practical significance lies in improving the quality of sorting potatoes and fruits and vegetables while eliminating the influence of the human factor by introducing a machine with a digital system for identifying contaminated biological objects and developing recommendations for post-harvest work when breeding new varieties and hybrids [5].
[1] P. Azizi, N.S. Dehkordi, R. Farhadi, Design, construction and evaluation of potato digger with rotary blade, Cercet. Agron. Mold. 2014, 47, pp.5-13.
[2] Abd. El-Rahman and M. Magda, Development and performance evaluation of a simple grading machine suitable for onion sets, Soil. Sci. and Agric. Eng. Mansoura Univ. 2014, 2, 213-226. p.9768.
[3] N.V. Byshov, S.N. Borychev, N.N. Yakutin, D.V. Kalmykov, N.V. Simonova, On the interaction of the tuberous layer with the digger's working organs, Bull. Ryazan State Agrotechnological Univ. P.A. Kostycheva, 2018, 40, pp.161-167.
[4] N.V. Byshov, N.N. Yakutin, R.Y. Koveshnikov, V.V. Rodionov, N.V. Serzhantov, P.S. Smirnov, Modernization of the KST-1.4 digger, Bull. Ryazan State Agrotechnological Univ. P.A. Kostycheva 2016, 30, pp.75-78.
[5] S. Bachche, K. Oka, Design, Modeling and Performance Testing of End-Effector for Sweet Pepper Harvesting Robot Hand, J. Robot. Mechatron, 2013, 25, pp.705-717.
