Mehitamata õhusõidukiga aerofotogramm-meetrilise mõõdistamise kõrgusliku täpsuse hindamine teedeehituse näitel
Laen...
Kuupäev
2020
Kättesaadav alates
26.11.2020
Autorid
Ajakirja pealkiri
Ajakirja ISSN
Köite pealkiri
Kirjastaja
Eesti Maaülikool
Abstrakt
Uurimustöö eesmärgiks oli uurida integreeritud RTK mooduliga mehitamata õhusõidukiga
aerofotogramm-meetrilise mõõdistuse kõrguslikku täpsust teedeehituse näitel. Võrreldi
erinevate trajektooridega lennuplaane, lennukõrguse ja tähiste 3D-punktipilve loomiseks
kasutatud tähiste arvu mõju 3D-punktipilve kõrguslikule täpsusele.
Uurimistöö objektiks oli tugimaantee nr 90 Põlva-Karisilla Rosma-Soohara lõigu
rekonstrueerimine. Andmete kogumine toimus 2019. aasta juulis ja augustis. Mõõdistati tee
kompleksstabiliseeritud kiht ja asfaltkatte pealmine kiht. Aerofotogramm-meetriliste lendude
sooritamiseks kasutati mehitamata õhusõidukit DJI Phantom 4 RTK. Mõlema kihi
mõõdistamiseks sooritati kolm fotogramm-meetrilist lendu: 50m kõrguselt ristuvate
trajektooridega lennuplaaniga, 50m kõrguselt lineaarse trajektooriga lennuplaaniga ja 40m
kõrguselt lineaarse trajektooriga lennuplaaniga. Saadud tulemusi võrreldi OÜ Raxoest-i poolt
teostatud teostusmõõdistustega, kus viimane võeti õigeks. Mõõdistuse kõrgusliku täpsuse
hindamiseks arvutati keskmised ruutvead (krv) ja asfaltkatte kihi paksus tee teljel. Kompleksstabiliseeritud kihi aerofotogramm-meetrilisel mõõdistamisel saavutati parimad
tulemused lennates 40m kõrguselt lineaarse trajektooriga lennuplaaniga ja 3D-punktipilve
loomisel kasutati ühte tähist (krv tee vasakus servas 0.015m, teljel 0.012m ja paremas servas
0.014m). Asfaltkatte pealmise kihi aerofotogramm-meetrilisel mõõdistamisel saavutati
parimad tulemused lennates 40m kõrguselt lineaarse trajektooriga lennuplaaniga ja 3Dpuntkipilve loomisel kasutati ühte tähist (krv tee vasakus servas 0.028m, teljel 0.016m ja
paremas servas 0.029m). Tähisteta 3D-punktipilve loomisel saavutati parim kõrguslik täpsus
kasutades ristuvate trajektooridega lennuplaani. Asfaltkatte paksuse arvutamisel saadi
parimad tulemused madalamalt lennates ja kasutades 3D-punktipilve loomisel kolme tähist
aga tähisteta saavutati kõige paremad tulemused kasutades ristuvate trajektooridega
lennuplaani.
Uurimistöö tulemusena selgus, et ristuvate trajektooridega lennuplaaniga lennates saavutati
tähisteta 3D-punktpilvest kõige täpsemad tulemused ja tähiseid kasutades saavutati
madalamalt lennates kõrgem 3D-punktipilve kõrguslik täpsus.
The aim of this Bachelor’s thesis is to evaluate vertical accuracy of aerial surveying using an unmanned aerial vehicle with integrated RTK module. Different flight plans, flight heights and ground control points effect on 3D-pointcloud vertical accuracy. The object of this study was reconstruction of basic road number 90 Põlva-Karisilla RosmaSoohara segment. Data was collected in August and July of 2019. Aerial survey was conducted using a DJI Phantom 4 RTK unmanned aerial vehicle. Collected data was then processed and compared with as-built survey conducted by OÜ Raxoest. Vertical accuracy was measured using root mean square error (rmse) and asphalt layer thickness in the middle of the road. The best results of the complex stabilized layer aerial survey were from a 3D-pointcloud made using one ground control point and flown with linear flight plan at 40m (root mean square error at left side of the road was 0.015m, middle of the road 0.012m and right side of the road 0.014m). The best results for the top layer of asphalt were also from a 3D-pointcloud made using one ground control point and flown with linear flight plan (root mean square error at left side of the road was 0.028m, middle of the road 0.016m and right side of the road 0.029m). The best results without ground control points were from using 3D cross flight plan. The same was for asphalt layer thickness. Where the best results were achieved while flying at lower altitude and with three ground control points but when no ground control points were used then the best results were achieved using cross flight plan. Results of the research show that, 3D cross flight plan is the most accurate when no ground control points are used but when ground control points are used then flying at lower altitude produces better results.
The aim of this Bachelor’s thesis is to evaluate vertical accuracy of aerial surveying using an unmanned aerial vehicle with integrated RTK module. Different flight plans, flight heights and ground control points effect on 3D-pointcloud vertical accuracy. The object of this study was reconstruction of basic road number 90 Põlva-Karisilla RosmaSoohara segment. Data was collected in August and July of 2019. Aerial survey was conducted using a DJI Phantom 4 RTK unmanned aerial vehicle. Collected data was then processed and compared with as-built survey conducted by OÜ Raxoest. Vertical accuracy was measured using root mean square error (rmse) and asphalt layer thickness in the middle of the road. The best results of the complex stabilized layer aerial survey were from a 3D-pointcloud made using one ground control point and flown with linear flight plan at 40m (root mean square error at left side of the road was 0.015m, middle of the road 0.012m and right side of the road 0.014m). The best results for the top layer of asphalt were also from a 3D-pointcloud made using one ground control point and flown with linear flight plan (root mean square error at left side of the road was 0.028m, middle of the road 0.016m and right side of the road 0.029m). The best results without ground control points were from using 3D cross flight plan. The same was for asphalt layer thickness. Where the best results were achieved while flying at lower altitude and with three ground control points but when no ground control points were used then the best results were achieved using cross flight plan. Results of the research show that, 3D cross flight plan is the most accurate when no ground control points are used but when ground control points are used then flying at lower altitude produces better results.
Kirjeldus
Bakalaureusetöö
Geodeesia ja maakorralduse õppekaval
Märksõnad
bakalaureusetööd, fotogramm-meetria, aerofotograafia, droonid