Ocena okroglega lesa na kamionskih cestah s pomočjo digitalne fotogrametrije

Lotrič, Katja (avtor), Kobal, Milan (mentor)

Magistrsko delo je bilo zastavljeno na podlagi možnosti uporabe fotogrametrije za izračun volumna okroglega lesa. Podatki so bili pridobljeni iz terenskih posnetkov skladovnic, ki so bile označene s štirimi kontrolnimi točkami. Skupno je bilo posnetih 37 skladovnic, ki so se delile v tri razširjene debelinske razrede. Sortimente v skladovnicah smo predhodno ročno izmerili ter označili z zaporednimi številkami. Dolžina posameznega sortimenta je bila 8,25 m. Volumen smo izračunali po Newtonovem obrazcu. Zajete posnetke smo obdelali z računalniškimi programi Pix4D, CloudCompare, ArcMap in Microsoft Excel. S prvim programom smo naredili oblak točk iz zajetih posnetkov. S poznavanjem natančne razdalje med kontrolnimi točkami smo oblak točk raztegnili v metrični sistem. S preostalimi programi smo izračunali volumen oblaka točk, ki je predstavljal skladovnice. Ugotovili smo, da sta dejanski volumen lesa v skladovnici in volumen rastrskega modela skladavnice v odvisnosti, ki jo najbolje opiše potenčna funkcija. Ugotovili smo, da napovedan volumen od dejanskega odstopa od -2,5 m3 do 1,5 m3. Povprečna napaka našega modela znaša 5,7 % (RMSE).

• Jezik: Slovenski jezik
• Ključne besede: fotogrametrija, oblak točk, volumen, skladovnica, ločljivost
• Vrsta gradiva: Magistrsko delo/naloga
• Organizacija: BF - Biotehniška fakulteta
• Leto izida: 2019
• PID: 20.500.12556/RUL-106368
• COBISS.SI-ID: 5322918
• Datum objave v RUL: 20.02.2019

Sekundarni jezik

• Jezik: Neznan jezik
• Naslov: Estimating roadside roundwood with digital photogrammetry

Izvleček

The master’s thesis explored photogrammetry as a potential method for calculation of roundwood volume. The data was acquired from field images of stacks which were marked with four control points. A total of 37 stacks divided into three diameter classes were recorded. Prior to recording, the logs in the stacks were manually measured and numbered. Log length was 8.25 m. Volume was calculated using Newton’s formula. The recorded images were processed with Pix4D, CloudCompare, ArcMap and Microsoft Excel software. Pix4D was used to create a point cloud from the acquired images. After measuring the exact distance between control points, we translated the point cloud into the metric system. The other software was used to calculate the volume of the point cloud representing stacks. We discovered that the correlation between the actual volume of wood in the stack and the volume of the raster model is best described with the power function. Forecast volume deviated from actual volume by -2,5 m3 to 1,5 m3. The average error rate of our model is 5.7% (RMSE).

• Ključne besede: photogrammetry, point cloud, volume, stacked, resolution