20.500.12556/RUL-140540 Detekcija terasiranih pokrajin kot semantična segmentacija digitalnega modela višin Detection of terraced landscapes as a semantic segmentation of a digital elevation model Slovenska pokrajina je široka, raznolika in prebivalstvu nudi tako kulturne kot tudi kmetijske funkcionalnosti. Temu služijo tudi terasirane pokrajine, ki so svetovno prepoznane kot pomembni naravni in kulturni element naše okolice. Ker Slovenija še nima ustreznega kriterija za razpoznavanje teh teras, v diplomskem delu preučujemo možnost načina klasificiranje takih pobočij s pomočjo semantične segmentacije z globokimi konvolucijskimi nevronskimi mrežami. Za detekcijo smo uporabili javno dostopne podatke LIDAR digitalnega modela višin, na katere smo aplicirali enostaven globoki klasifikacijski model z arhitekturo U-Net. Referenčni podatki so bili anotirani s strani strokovnjakov, vendar so anotacije nenatančne in pomanjkljive. Dosežen napovedni model smo naučili dobro locirati terasirane pokrajine. Doprinesel je tudi nova spoznanja o terasiranih pokrajinah in referenčnih podatkih, ki bodo v korist slovenskim geografom. Prikazali smo tudi rezultate, ki prikazujejo, da je, kljub šumnim podatkom, detekcija takih pokrajin možna z globokimi modeli in predstavlja odskočno desko za nadaljnje raziskave v smeri izboljšanja njihove zaznave. The Slovenian landscape is wide, diverse and above all offers cultural as well as agricultural functionalities to the population. Terraced landscapes contain an abundance of cultural as well as functional importance and must be identified and preserved. Since Slovenia does not yet have an adequate criterion for the identification of these terraces, we present in a way of classifying them by means of semantic segmentation with deep convolution neural network. For detection, we used the publicly available digital elevation model (LIDAR), where a simple detection model with U-Net architecture was trained. The reference data have been annotated by experts, but the annotations are inaccurate and flawed. The model managed to learned to locate terraced landscapes, which offered insights about terraced landscapes and reference data that will benefit Slovenian geographers. The results also show that, despite the noise data, detection of such landscapes is possible and represents a stepping stone for further research towards improving their detection. terase digitalni model višin semantična segmentacija umetna inteligenca strojno učenje LIDAR terraces digital elevation model semantic segmentation artificial inteligents machine learning LIDAR true false false Slovenski jezik Angleški jezik Diplomsko delo/naloga 2022-09-15 14:20:00 2022-09-15 14:20:03 2022-10-03 05:38:00 0000-00-00 00:00:00 2022 0 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 35068 123820035 Glusic_Anze_-_Detekcija_terasiranih_pokrajin_kot_semanticna_segmentacija_digitalnega_modela_visi.pdf Glusic_Anze_-_Detekcija_terasiranih_pokrajin_kot_semanticna_segmentacija_digitalnega_modela_visi.pdf 1 6B328FF65DBB221C7F5DDDDB012C6B35 ac2590eb3c1e89608e93da0a0f5b7ef68e83a419385bd8b6e9648b78fdb1e002 a9c39e84-34f0-11ed-92af-00155dcfd717 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=161293 Fakulteta za računalništvo in informatiko 0 0 0