Iskanje po katalogu digitalne knjižnice

Opis: This study focuses on the segmentation part in the development of a potato-sorting system that utilizes camera input for the segmentation and classification of potatoes. The key challenge addressed is the need for efficient segmentation to allow the sorter to handle a higher volume of potatoes simultaneously. To achieve this, the study employs a region-based convolutional neural network (R-CNN) approach for the segmentation task, while trying to achieve more precise segmentation than with classic CNN-based object detectors. Specifically, Mask R-CNN is implemented and evaluated based on its performance with different parameters in order to achieve the best segmentation results. The implementation and methodologies used are thoroughly detailed in this work. The findings reveal that Mask R-CNN models can be utilized in the production process of potato sorting and can improve the process.
Ključne besede: image segmentation, potato sorting, neural network, mask RCNN, object detection, production process, machine learning, AI
Objavljeno v DKUM: 27.03.2025; Ogledov: 0; Prenosov: 9
Celotno besedilo (5,97 MB)
Gradivo ima več datotek! Več...

Opis: Large binary images are used in many modern applications of image processing. For instance, they serve as inputs or target masks for training machine learning (ML) models in computer vision and image segmentation. Storing large binary images in limited memory and loading them repeatedly on demand, which is common in ML, calls for efficient image encoding and decoding mechanisms. In the paper, we propose an encoding scheme for efficient compressed storage of large binary images based on chain codes, and introduce a new single-pass algorithm for fast parallel reconstruction of raster images from the encoded representation. We use three large real-life binary masks to test the efficiency of the proposed method, which were derived from vector layers of single-class objects – a building cadaster, a woody vegetation landscape feature map, and a road network map. We show that the masks encoded by the proposed method require significantly less storage space than standard lossless compression formats. We further compared the proposed method for mask reconstruction from chain codes with a recent state-of-the-art algorithm, and achieved between and faster reconstruction on test data
Ključne besede: binary mask, machine learning, chain code, binary encoding, bitmap reconstruction
Objavljeno v DKUM: 29.01.2025; Ogledov: 0; Prenosov: 7
Celotno besedilo (1,45 MB)

Opis: The production of personal protective equipment (PPE) has increased dramatically in recent years, not only because of the pandemic, but also because of stricter legislation in the field of Employee Protection. The increasing use of PPE, including disposable surgical masks (DSMs), is putting additional pressure on waste collectors. For this reason, it is necessary to find high-quality solutions for this type of waste. Mechanical recycling is still the most common type of recycling, but the recyclates are often classified as low-grade materials. For this reason, a detailed analysis of the recyclates is necessary. These data will help us to improve the properties and find the right end application that will increase the value of the materials. This work represents an extended analysis of the recyclates obtained from DSMs, manufactured from different polymers. Using surface and morphology tests, we have gained insights into the distribution of different polymers in polymer blends and their effects on mechanical and surface properties. It was found that the addition of ear loop material to the PP melt makes the material tougher. In the polymer blends obtained, PP and PA 6 form the surface (affects surface properties), while PU and PET are distributed mainly inside the injection-molded samples.
Ključne besede: mechanical recycling, disposable surgical mask, morphology, surface properties, mechanical properties, nonwoven materials, PPE
Objavljeno v DKUM: 09.04.2024; Ogledov: 232; Prenosov: 22
Celotno besedilo (8,08 MB)
Gradivo ima več datotek! Več...

Opis: Pričujoče magistrsko delo že v naslovu opozarja na eksistencialni in identitetni problem romanesknega subjekta v izbranih Žabotovih romanih Pastorala, Stari pil, Volčje noči in Sukub. Vsi, razen zadnjega, ki je Žabotov t. i. »urbani« prvenec, so postavljeni v morbidno močvirsko okolje, prostor brezmejne ravnice, ki s svojimi mitološkimi in polpoganskimi spremljevalci že ob prvem stiku svoje popotnike in večne iskalce napolni s patološkim hrepenenjem po Bližnjem in absolutni resnici. Blodnjava motnja postaja protagonistov bivanjski označevalec, ki ga vodi skozi temne, plesnive in vlažne hodnike zamejenih prostorov kolektivnega, s težkim zrakom prepojene beznice, skozi brezbožne ustanove božjega, vse do podstrešij stolpnic, kjer bi bilo mogoče najti še kako varno zatočišče pred »zadušljivimi« načeli kolektivnega. Paranoja postaja vzvod subjektovega večnega bega pred zasledovalnim principom fatalno erotičnega in kolektivnega, pri čemer oba pridobivata destruktivne razsežnosti. Rešitev eksistenčnega položaja in razkrajajoče se identitete je za protagonista možna le s pobegi v trans-resnico ali z nadevanjem »mask« po dolžnosti. Hrepenenje po idealni ljubezni do Bližnjega se z vsakim nadaljnjim korakom izkaže kot prazen up. Junakovi napori po rešitvi se kljub zasledovalnim in omejevalnim principom v celoti nikoli ne končajo. V eksistenčnih prehodih se protagonist psihično in fizično znajde »na meji« in šele na zadnji bivanjski postaji jo za hip prestopi. Tu ugleda pravega Bližnjega in spozna absolutno Resnico. Pot Žabotovega romanesknega subjekta se nikoli ne konča. Na svoji zadnji postaji se vselej vrne na začetek.
Ključne besede: Vlado Žabot, subjekt, romaneskni subjekt, protagonist, eksistenca, identiteta, razkroj, eksistenčni prehodi, nadevanje »mask«, Bližnji, Drugi, kolektivni princip, erotični princip, »ljubiti po dolžnosti«, absolutna Resnica
Objavljeno v DKUM: 05.10.2016; Ogledov: 1663; Prenosov: 186
Celotno besedilo (2,08 MB)