magistrsko delo
Blaž Krpan (Author), Božidar Potočnik (Mentor)

Abstract

Pšenica je ena izmed najbolj priljubljenih žitaric v prehrambni industriji, zato je nadzor kakovosti pšenice ključnega pomena. V okviru magistrske naloge smo pokazali, da je mogoče proces nadzora kakovosti pšenice avtomatizirati in izboljšati v primerjavi z današnjim pristopom. Reševanja problema smo se lotili z računalniško analizo digitalnih slik in spektralnih odzivov zrn ter uporabo algoritmov strojnega učenja. Posebno pozornost smo posvetili združevanju obeh tipov podatkov. Prišli smo do zaključka, da lahko z natančnostjo 90 % ali več razvrstimo zrna zdrave pšenice in ostalih 8 razredov, ki predstavljajo poškodovana zrna in druga žita. Za potrditev rezultatov in izboljšanje robustnosti algoritma predlagamo še dodatne eksperimente.

Keywords

nadzor kakovosti;pšenica;strojno učenje;razvrščanje;obdelava slik;

Data

Language: Slovenian
Year of publishing:
Typology: 2.09 - Master's Thesis
Organization: UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher: B. Krpan
UDC: 004.93'1:633.11(043.2)
COBISS: 19317014 Link will open in a new window
Views: 878
Downloads: 74
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: Wheat grains quality control with advanced methods for image processing and pattern recognition
Secondary abstract: Wheat is one of the most produced cereals for food industry worldwide. Therefore quality control of wheat is an essential part of the supply chain. In the thesis, it has been shown that it is possible to automatize and improve the wheat quality inspection procedures used today. To solve the problem digital images and spectral responses of the grains have been analysed in combination with the machine learning algorithms. Additional effort has been spent on combining the two types of data. Conclusion has been made that it is possible to distinguish between the healthy wheat and the other 8 classes, which represent damaged grains and other cereals, with the accuracy of 90% or more. In order to confirm the results and improve the robustness of the developed algorithm further experiments shall be made.
Secondary keywords: quality inspection;wheat;machine learning;classification;image processing;
URN: URN:SI:UM:
Type (COBISS): Master's thesis/paper
Thesis comment: Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Računalništvo in informacijske tehnologije
Pages: VII, 62 f.
ID: 9055844