TTK Links: Home Page | TTK Moodle | ÕIS
TTK/UAS Repository For Digital Teaching and Learning Resources

Laserkiire langemisnurga mõju mõõdistuspunktide täpsusele ja 3D punktipilve usaldatavus

Paatsi, Simo (2018) Laserkiire langemisnurga mõju mõõdistuspunktide täpsusele ja 3D punktipilve usaldatavus. [thesis] [en] The Influence of the Angle of Incidence upon the Quality of the Point Cloud and the Reliability of the 3D Point Cloud.

[img] PDF - Published Version
Restricted to Registered users only

Download (2MB)
[img] Image (JPEG) (lihtlitsents) - Other
Restricted to Registered users only

Download (373kB)

Abstract

Lõputöö eesmärgiks oli välja selgitada laserkiire langemisnurga mõju mõõdistuspunktide täpsusele ning uurida 3D punktipilve usaldatavust ehk selle õigsust. Lõputöö käigus tuvastati selge muutus punktipilves langemisnurga suurenemisel, mis kinnitab kirjanduses levivat väidet, et kui skaneerimisel on langemisnurk üle 60. kraadi, langeb punktipilve kvaliteet. Käesoleva lõputööga anti ülevaade laserskaneerimise vigade analüüsist ja nende tekkest. Võrreldi nelja laserskannerit homogeensetes tingimustes. Katsed viidi läbi Tallinna Tehnikakõrgkooli keldrikorrusel. Katse mõõdistusobjektiks oli mattvalge tahvel, millel asendinurka muudeti fikseeritud ringmalli abil iga 10. kraadi järel. Antud töös kasutati järgnevaid laserskannereid: Leica P40, Leica HDS6200, Leica C10 ja Trimble TX5. Punktipilve õigsuse kontrollimisel saadi kõige tõepärasem tulemus Leica C10 puhul (-1 mm). Kui võrrelda kõikide kasutatud skannerite ja elektrontahhümeetri kauguse mõõtmise arvutatud kombineeritud mõõtemääramatust arvutatud hälvetega, siis kõik hälbed jäävad mõõtemääramatuse piiresse ehk instrumentide spetsifikatsioonid peavad paika. Katsete tulemustest selgus, et mida suurem on langemisnurk, seda vähem punkte on mõõdetud pinnal. Samuti selgus, et ka mõõdetud pinna geomeetrilised mõõtmed muutuvad märkimiseväärselt kui langemisnurk on suurem kui 40 kraadi. Kõige suurem langemisnurga moonutus pinna geomeetriliste mõõtmete suhtes ja punktide paiknemisel pinnal on katses kasutatud skanneritest Leica HDS6200 ja Trimble TX5 puhul. Mõlemad kasutavad mõõtmiseks faasinihke meetodit. Leica C10-nel ja Leica P40-nel esinesid väiksemad hälbed pinna mõõtmetes. Samuti nende punktide paiknemine pinnal oli korrapärasem kui Leica HDS6200-l ja Trimble TX5-l. Mida suurem oli langemisnurk, seda nõrgem on tagasipeegeldunud signaal, mille tõttu oli ka punkte järjest vähem langemisnurga suurenemisel. Antud lõputöö tulemusest võib järeldada, et mida suurem on langemisnurk seda suuremad on vead ja ebakõlad punktipilves. Langemisnurk peaks jääma alla 40. kraadi, siis saab veel skannerist tulevaid andmeid usaldada. Kui aga langemisnurk läheb üle 40. kraadi tasub kaaluda, kas neid andmeid saab kasutada, ja kui usaldusväärsed nad on.

Abstract [en]

The heading of the following thesis is The Influence Of The Angle Of Incidence Upon The Quality Of The Point Cloud And The Reliability Of The 3D Point Cloud. The investigation was carried out using terrestrial laser scanners with different distance-measuring principles. The thesis revealed a clear connection between an increase in the angle of incidence and the quality of the point cloud. When the angle of incidence exceeds an arc of approximately sixty degrees, the quality of the point cloud decreases. The latter, in general, confirms the much-debated statement that has been made in the available literature. The thesis is divided into two parts, the first being an overview of the uncertainty sources of laser scanning and the second covering a description of the experiment that has been conducted here. The description of the experiment also includes the results. The experiments were carried out at the Tallinn University of Applied Sciences (the TTK UAS). The measurements took place during a single day under homogeneous conditions. For the experiment, the object being measured was an opaque white board. The orientation angle of the board and therefore also the angle of incidence was changed at increments of ten degrees using a round protractor attached to the tribrach. The laser scanners being used for the process were the following: Leica HDS6200 and Trimble TX5 phase shift scanners, a time of flight Leica C10 laser scanner, and a waveform digitising Leica P40 laser scanner. In order to analyse the point cloud data, the equation for the plane was fitted into the data using the least squares method. The reliability of the 3D point cloud was tested because this would make it possible to discover the point locations from the real board position, either above or behind. For this, the board’s reference distance was determined using the Trimble S9 total station. The reference distance was determined using the true value of eight distances. The distance between the board and the total station was considered as being the probable figure. The laser scanner and total stations both have their own uncertainties in terms of their measurements, so in order to be able to discover the actual true distance, the uncertainty of the measurements themselves must be considered. The Leica C10 came up with the most realistic results (-1mm). When comparing the calculated combined uncertainty of measurements with the calculated deviations in terms of the scanner and the total station, all tolerances can be seen to remain within the uncertainty range. This means that the specification for the instruments is correct. The results were studied in terms of the influence of the angle of incidence on the quality of the point cloud by basing efforts upon earlier work by Lichti (2007) and Soudarissanane (2011). The results indicated that the statements being debated in the literature are correct. The standard deviation changes more drastically when the angle of incidence increases over sixty degrees. The total values of minimum and maximum residues shows that, thanks to the higher angle of incidence, the point cloud changes twice (Leica P40). In addition, the results indicated that the larger is the angle of incidence the less points are measured on the surface. It also turned out to be the case that the geometric dimensions of the measured surface changed dramatically when the angle of incidence exceeded only forty degrees. The results revealed that the largest distortions in the geometric dimensions of the board and the locations of the survey points with respect to the surface of the board itself were experienced when using the Leica HDS6200 and the Trimble TX5. In addition, the location of the points on the surface were more regular in the case of the Leica C10 and the Leica P40 than they were for the Leica HDS6200 and the Trimble TX5. When the angle of incidence exceeds a certain threshold, the reflected signal is not recorded by the scanner and, as a result, there are less points on the surface of the object. To conclude, the results revealed that the larger is the angle of incidence the more it influences the quality of the obtained point cloud. It is suggested that, for rigorous geodetic surveys, the angle of incidence should not exceed forty degrees, especially in the case of phase shift scanners. In general, one can trust the data that has been gathered in scanning under the range of the angle of incidence between zero to forty degrees. If the angle of incidence exceeds forty degrees, than one should consider whether this data should be used at all and whether the data is reliable.

Item Type: thesis
Advisor: Tarvo Mill
Subjects: Construction > Applied Geodesy
Divisions: Institute of Construction > Applied Geodesy
Depositing User: Simo Paatsi
Date Deposited: 06 Jun 2018 06:08
Last Modified: 27 Aug 2019 07:56
URI: http://eprints.tktk.ee/id/eprint/3843

Actions (login required)

View Item View Item