izpis_h1_title_alt

Analiza možnosti uporabe samodejne klasifikacije podobe za vzdrževanje prostorskih podatkovnih zbirk
Mesner, Nika (Author), Oštir, Krištof (Mentor) More about this mentor... This link opens in a new window

.pdfPDF - Presentation file, Download (3,80 MB)
MD5: 7FBF599C97DA28521176D84E61D57159
URLURL - Presentation file, Visit http://drugg.fgg.uni-lj.si/5518/ This link opens in a new window

Abstract
Vzdrževanje prostorskih podatkov s fotointerpretacijo je časovno in kadrovsko ter zato tudi finančno zelo zahtevno. Težnje upravljavcev podatkov po avtomatizaciji postopka so zelo velike. Kljub številnim raziskavam na tem področju se samodejne metode vzdrževanja v praksi redko uporabljajo. V magistrski nalogi sem preučila možnosti uporabe objektno usmerjene analize za določitev pozidanih površin na podlagi podatkov daljinskega zaznavanja. Izvedena je analiza vpliva prostorske, spektralne in časovne ločljivosti ter filtriranja in transformacije podobe na kakovost obeh faz objektno usmerjene analize, segmentacije in objektne klasifikacije. Na podlagi rezultatov analize je predlagana optimalna izbira ločljivosti podatkov in predobdelave posnetka za razpoznavo pozidanih površin. Podan je pregled podatkovnih zbirk v Sloveniji, v katerih se vodijo podatki o stavbah, topografiji, pokrovnosti tal in rabi prostora. Ocenjena je tudi možnost uvedbe samodejnih postopkov v proces vzdrževanja. Na testnem območju so na podlagi ortofota in lidarskih podatkov določeni obrisi stavb in druge pozidane površine. Na podlagi enotnega sloja obrisov stavb so identificirane spremembe glede na evidentirano stanje v katastru stavb in dejanski rabi kmetijskih in gozdnih zemljišč z visoko stopnjo popolnosti (več kot 80 %). Bistvena prednost uvedbe samodejnih postopkov v proces vzdrževanja prostorskih podatkov je, da s samodejno identificiranimi spremembami usmerimo operaterja na zajem podatkov le na območja, kjer so spremembe.

Language:Slovenian
Keywords:objektno usmerjena analiza, segmentacija, vzdrževanje prostorskih podatkov, pozidanost tal, iskanje sprememb
Work type:Master's thesis (m2)
Tipology:2.09 - Master's Thesis
Organization:FGG - Faculty of Civil and Geodetic Engineering
Year:2016
Publisher:[N. Mesner]
Place:Ljubljana
UDC:528.8:659.2:91(043.3)
COBISS.SI-ID:7500897 This link opens in a new window
Views:1452
Downloads:575
Metadata:XML RDF-CHPDL DC-XML DC-RDF
 
Average score:(0 votes)
Your score:Voting is allowed only to logged in users.
:
Share:AddThis
AddThis uses cookies that require your consent. Edit consent...

Secondary language

Language:English
Title:Analysis of the possibility of using automatic image classification for the maintenance of spatial databases
Abstract:
Maintenance of spatial data with photointerpretation is very demanding in terms of time and personnel, therefore tendencies of data owners for the automation of this process are very high. Despite extensive research on automatic extraction of spatial data, these methods are still not in practical use. In the thesis, I examined the possibility of using object-based image analysis for extraction of the built-up areas with remote sensing data. An analysis of the impact of spatial, spectral and temporal resolution, image filtering and transformation on the quality of object-based image analysis, has been carried out. Based on the results of the analysis the optimal resolution and image pre-processing has been proposed. An overview of buildings, topography, land cover and land use spatial databases in Slovenia is given. The possibility of introducing automated processes in the maintenance process of specific spatial data has been assessed. Object based image analysis has been used for extraction of buildings contours and other built-up areas. This single data layer of buildings and other built-up areas is used for change detection in comparison to registered data in different databases. Completeness of detected changes is very high (more than 80%). The major advantage of proposed automated change detection in the spatial data maintenance process is that the automatically identified changes direct the operator to capture only data on the areas where change exists and so significantly reduces the amount of work.

Keywords:object based image analysis, segmentation, spatial data maintenance, built-up area, change detection

Similar documents

Similar works from RUL:
Similar works from other Slovenian collections:

Comments

Leave comment

You have to log in to leave a comment.

Comments (0)
0 - 0 / 0
 
There are no comments!

Back