magistrsko delo
Marko Brodarič (Author), Janez Perš (Mentor), Marija Ivanovska (Co-mentor)

Abstract

Proso je žito, ki je v zadnjih letih vse bolj razširjeno v prehrambeni industriji, saj ima številne ugodne učinke na zdravje ljudi, obenem pa s svojo robustno rastjo omogoča ekološki način pridelave. V postopkih obdelave so prosena zrna deležna večih stopenj čiščenja. Največ izgub od teh proizvede zadnja stopnja, ki ločuje neoluščena zrna prosa od prosene kaše, saj je le-te težko ločiti s sistemi sit in odsesav, ročno razvrščanje pa je časovno izredno počasno. V sklopu te magistrske naloge smo naslovili ta problem tako, da smo razvili mehanski sistem, ki nadzorovano upravlja z zrni, zajema slike ter izloča tujke in programski del, ki zajete slike obdela ter klasificira bodisi med dobra zrna bodisi med tujke. Le-tega smo zasnovali na podlagi enorazrednih metod za zaznavanje anomalij. Testirali smo obstoječi metodi PaDiM in CS-Flow. Z namenom izboljšanja uspešnosti delovanja omenjenih algoritmov pa smo razvili tudi svojo metodo. Tako smo pri 1 % neoluščenih zrn v končnem produktu dosegli manj kot 5 % zrn prosene kaše med odstranjenimi anomalijami. Tekom razvoja sistema je nastala tudi podatkovna zbirka zrn prosene kaše ter neoluščenega prosa.

Keywords

optično sortiranje;prosena kaša;računalniški vid;globoke nevronske mreže;enorazredno zaznavanje anomalij;CS-Flow;PaDiM;magisteriji;

Data

Language: Slovenian
Year of publishing:
Typology: 2.09 - Master's Thesis
Organization: UL FE - Faculty of Electrical Engineering
Publisher: [M. Brodarič]
UDC: 004.93:633.17(043.3)
COBISS: 167987459 Link will open in a new window
Views: 52
Downloads: 19
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: Automated separation of unhulled proso grains from proso millet
Secondary abstract: Proso millet is a grain crop that has been increasingly popular in the food industry in recent years due to its numerous health benefits, while also enabling eco-friendly cultivation with its robust growth. During the processing, proso grains go through several stages of cleaning. The most significant losses occur in the final stage, which separates unhulled proso grains from the hulled proso millet, as it is challenging to distinguish them using sieves and suction systems, while manual sorting is extremely time-consuming. We addressed this problem by developing a mechanical system that handles grains in a controlled manner and removes foreign particles, based on visual analysis of images. Image data is collected with an embedded computer vision system, specially designed for quality inspection. Images of grains are processed with one-class learning anomaly datection algorithms, which classify the grains as either good or bad. We first tested state-of-the-art methods PaDiM and CS-Flow. We further improved anomaly detection rates by combining the advantages of both approaches. As a result, we achieved less than 5 % hulled proso millet among the removed anomalies at 1 % unpeeled proso grains in the final product. Throughout the system's development, a dataset of hulled proso millet and unhulled proso grains was created.
Secondary keywords: optical sorting;proso millet;computer vision;deep neural networks;one-class anomaly detection;CS-Flow;PaDiM;
Type (COBISS): Master's thesis/paper
Study programme: 1000316
Embargo end date (OpenAIRE): 1970-01-01
Thesis comment: Univ. v Ljubljani, Fak. za elektrotehniko
Pages: XX, 54 str.
ID: 19963924