Satelite bidezko irudien segmentazio semantikoa
| dc.contributor.advisorTFE | Pérez Goya, Unai | |
| dc.contributor.affiliation | Escuela Técnica Superior de Ingeniería Agronómica y Biociencias | es_ES |
| dc.contributor.affiliation | Nekazaritzako Ingeniaritzako eta Biozientzietako Goi Mailako Eskola Teknikoa | eu |
| dc.contributor.author | Macicior Michelena, Jaione | |
| dc.date.accessioned | 2024-08-29T17:32:11Z | |
| dc.date.available | 2024-08-29T17:32:11Z | |
| dc.date.issued | 2024 | |
| dc.date.updated | 2024-08-27T09:12:35Z | |
| dc.description.abstract | Azken urteotan, droneak tresna berritzaile gisa agertu dira nekazaritza-ingurumeneko teledetekzioan. Dronetan muntatutako kamerek urruneko sentsore pasibo gisa jokatzen dute eta kalitatezko irudiak modu erraz eta eraginkorrean lortzea ahalbidetu dute. Satelite bidez lortutako irudiek ere funtsezko tresnak dira teledetekzioan, Lurrari buruzko informazio sorta zabala ematen baitute. Gainera, dagoeneko dauden hainbat webgune nahiz prgramei esker, informazio asko eskuragarri dago. Proiektu honetan irudi multiespektro satelitalak erabiliko dira kalitatezko datuak lortzeko eta segmentazioa modu erraz eta eraginkorrean egiteko. Prozesu honen helburua drone bidez lortutako edozein irudi segmentatzeko gai den sare neuronal bat garatzea da, horrela aldez aurretik eskuz etiketatzeko lana saihestuz. Bertze modu batean esanda, eskuragarri dauden kalitate handiko irudiak erabiliko dira ondoren lortutako irudiei segmentazio semantikoa aplikatzeko, eskuzko lana ahalik eta gehien minimizatuz. | eu |
| dc.description.abstract | In recent years, drones have emerged as innovative tools in agricultural and environmental remote sensing. Cameras mounted on drones act as passive remote sensors, enabling the acquisition of high-quality images easily and efficiently. Satellite-acquired images are also essential tools in remote sensing, providing a wide range of information about the Earth. Moreover, a wealth of information is already available thanks to various existing websites and programs. In this project, multispectral satellite images will be used to obtain high-quality data and perform segmentation in an easy and efficient manner. The goal of this process is to develop a neural network capable of segmenting any image obtained by drones, thereby avoiding the need for manual labeling in advance. In other words, high-quality available images will be used to apply semantic segmentation to subsequently acquired images, minimizing manual labor as much as possible. | en |
| dc.description.degree | Graduado o Graduada en Ciencia de Datos por la Universidad Pública de Navarra | es_ES |
| dc.description.degree | Datuen Zientzietako Graduatua Nafarroako Unibertsitate Publikoan | eu |
| dc.format.mimetype | application/pdf | en |
| dc.identifier.uri | https://academica-e.unavarra.es/handle/2454/51497 | |
| dc.language.iso | eus | |
| dc.rights.accessRights | info:eu-repo/semantics/openAccess | |
| dc.subject | Segmentazio semantikoa | eu |
| dc.subject | PyTorch | eu |
| dc.subject | DeepLabV3 | eu |
| dc.subject | Resnet50 | eu |
| dc.subject | Google Earth Engine | eu |
| dc.subject | Semantic segmentation | en |
| dc.subject | PyTorch | en |
| dc.subject | DeepLabV3 | en |
| dc.subject | Resnet50 | en |
| dc.subject | Google Earth Engine | en |
| dc.title | Satelite bidezko irudien segmentazio semantikoa | eu |
| dc.type | info:eu-repo/semantics/bachelorThesis | |
| dspace.entity.type | Publication | |
| relation.isAdvisorTFEOfPublication | 003960f8-aa6d-47ec-bec1-6bc59c3c8edb | |
| relation.isAdvisorTFEOfPublication.latestForDiscovery | 003960f8-aa6d-47ec-bec1-6bc59c3c8edb |