Jürgen giese technology mix in practice building documentation at bamberg cathedral

For a comprehensive documentation of a building that meets scientific requirements, the basic procedures – as is also usual on archaeological excavations – should be the drawing, verbal and photographic recording in equal measure and, depending on the question, should be supplemented by additional investigations [1]. Within the three basic procedures mentioned above, the recording of building geometries, which is usually referred to as construction measurement, occupies a special position due to the time and technical effort required for this purpose. However, this effort is justified, since the building survey is an indispensable basis for scientific investigations as well as for planning work that precedes, for example, maintenance, renovation and conversion measures.

At the University of Bamberg, therefore, since the establishment of the subject of building research and building history, building documentation, including building surveys, has had a high priority in teaching and research, where it plays a large part, for example, in the master's degree program in monument preservation. The building measurement is always seen as the result of the interplay of "measuring" and "measuring" and "Display comprehended. A geometrically correct and meaningful building measurement can only be achieved if both steps are seamlessly intertwined. All methods used for geometry recording and representation must be subordinated to the goal of correctly representing as much knowledge as possible about the construction and surfaces of the structures under investigation. For the execution of such work, the geodetically trained building researcher or the geodetically trained building researcher is required, both of whom can equally emerge from the above-mentioned course of studies [2].

In 2010, the project "Bamberg Cathedral Digital" was launched a three-year project has been started, in which the existing competences are used and the above mentioned claim is demonstrated by a prominent example [3]. The aim of the project is to provide a complete geometric survey of the Bamberg Cathedral, which shares its fate with many other important testimonies of architectural art: For the preservation of its building fabric and its cultural-historical testimonial value, researchers, planners and building practitioners require precise plan documents that depict the structure in a comprehensive manner and in accordance with the building's components. However, such a set of plans, which meets current requirements in terms of scope and detail, is currently available neither for the Bamberg Cathedral nor for many comparable buildings. The aforementioned project is intended to remedy this situation.

Nowadays, at the beginning of such a project, the choice of the surveying product is made, which can be selected from a wide range of hand drawings up to digital surface and volume models. Here, the digital line drawing was chosen as the final product for two reasons:

1. With the skills and techniques currently available in building authorities, planning offices and research institutes, this product can be further processed without any problems or losses. The digital plan set thus fulfills the criterion of being a planning instrument that can be updated from the beginning and that accompanies and depicts both research companies and building maintenance in the coming decades. For an object of the size of the Bamberg Cathedral in particular, this has a head start over analog line drawings, point clouds and three-dimensional digital models. 2. In accordance with the principle that a meaningful survey should depict the construction and surfaces of the examined structures correctly in terms of content and geometry, the survey product requires final inspection and usually also supplementation by the construction researcher. However, changes and additions are difficult to make in automatically generated products such as point clouds and surface models and would result in hybrid products that are difficult to handle. In line drawing, on the other hand, the geometric data obtained using a wide variety of techniques is converted into a uniform representation in which important elements for understanding the building can be worked out and observations that are often made in hidden places can be emphasized. Only in the line drawing is the building researcher in a position to pass on his observations and experiences made on the building directly to the user without much effort.

Although the digital line drawing always works on a scale of 1:1, a target scale must be defined in the next step, which decides on the accuracy of the measurements used and the detail of the representation to be achieved. For a long time, the scales of 1:20 and 1:2 have been used for recording architecture in a way that is appropriate to the building component. 1:25, which thus also form the guideline for this project. The measurement accuracy is thus fixed to a standard deviation of ±5 mm per point, which results from the representation and tapping accuracy of analog drawings of these scales. As is also usual for analog drawings, this standard deviation can be deviated from depending on the component accuracy. Less precisely manufactured components such as quarry stones require lower accuracies, whereas the manufacturing precision of some wood and metal components also allows for higher measurement accuracy.

On the detailing side, the principle adopted from the analog drawings of these scales is that all components, including the plaster edges and building ornamentation, are shown with outline and internal edges. However, the surface portrait of the depicted components, which is common in analog drawings, has been omitted, for example the tool marks of the natural stone, wood and plaster work. These would be reliable only by autopsy and with targeted grazing light on site, d.H. Usually from a scaffold, which of course is not available for the entire Bamberg Cathedral. Due to time constraints, the detailing of furnishings such as the organ, pews, and sculptures is also deviated from, which is optimized for an easily legible representation at a scale of 1:100 (Fig. 1).

Jürgen giese technology mix in practice building documentation at bamberg cathedral
Jürgen giese technology mix in practice building documentation at bamberg cathedral

Fig. 1 Bamberg Cathedral, section of the longitudinal section with view to the north. Evaluation of sculptures and equipment: J. Müller, evaluation lithotomy: G. Kröck, addition on site: A. Hager. Installation: Jürgen Giese

For the geometry capture, the technique of terrestrial laser scanning is used in a first step. A time-of-flight scanner is used especially in the exterior, since distances of more than 150 m have to be bridged, for example, for the detection of the towers, in order to let the measuring beam hit the facade surfaces at the steepest possible angle. In order to achieve the required measurement accuracy during the subsequent evaluation, the average point spacing is set at 3 to 5 mm [4]. In the interior, on the other hand, a phase comparison scanner is mainly used, which combines high measuring speed with ease of use. The range of the scanner, which is stated by the manufacturer to be 120 meters, is limited to 30 meters for this project, since an average point spacing of ≤5 mm can only be achieved up to this distance [5].

The scans are oriented using reference points whose coordinates are known in the fixed point system of the cathedral, which was determined by geodetic compensation calculation with maximum standard deviations of ±5 mm [6].

In addition to the scans, high-resolution digital photos are taken of all wall surfaces [7], as the photos provided by the scanner's internal cameras cannot be used for three reasons:

1. The required imaging scale of 2 to 3 mm per pixel cannot be achieved, especially at greater distances of the scanner from the object. 2. The separately produced photography can be used – for example by using a high tripod, a cherry picker and a photo drone (Fig. 2 and 3) – show components that lie in the scan shadow, i.E. For which no data is available in the scan. 3. The amount of light used for the photo can be specifically controlled by adding artificial light, especially indoors, which significantly improves the imaging quality compared to the scanner's own cameras.

This initial work on site takes up about 10 percent of the project's working time.

Jürgen giese technology mix in practice building documentation at bamberg cathedral

Fig. 2: Bamberg Cathedral, exterior inspection with the elevating platform for photo documentation. Photo: A. Hager

Jürgen giese technology mix in practice building documentation at bamberg cathedral

Fig. 3: Bamberg Cathedral. Flight of the facade of the southwest tower with photo drone. Photo: Jürgen Giese

In the next step, the oriented point clouds in combination with the external photos are converted into line drawings in the office. Various methods are available for this purpose. For all sculpturally protruding component edges, the evaluation exclusively in the point cloud has proven to be successful (Fig. 4), whereby the additional photos are usually viewed in parallel on a second screen in order to be able to better interpret the evaluated component. The use of only the point cloud allows to change the viewing direction on it in a satisfying speed and thus to apply evaluation strategies that are familiar from the tachymetry on site. These evaluation strategies are especially important when component edges or the correct projections of component edges have to be constructed which are poorly preserved or fuzzy defined. An indispensable prerequisite for the creation of a meaningful line drawing is the use of a full-fledged d.H. CAD system working in three-dimensional space with numerous design and drawing tools. The drawing tools that are – if at all – directly integrated into the evaluation software of the scanner manufacturers can be described as rudimentary at best and are unsuitable for complete measurement drawings.

Jürgen giese technology mix in practice building documentation at bamberg cathedral
Jürgen giese technology mix in practice building documentation at bamberg cathedral

Fig. 4 Bamberg Cathedral, Adam's Gate. Evaluation of sculptures in the point cloud. Left view, right top view of point cloud and line drawing. Evaluation J. Müller. Installation: Jürgen Giese

In a further step, the component boundaries lying on continuous surfaces are evaluated, e.G. The stone cut or plaster boundaries on wall surfaces. For this evaluation, it is crucial how well the scanner used was able to differentiate the reflectance values of the appropriate surfaces and how these intensity values, usually presented as gray levels, are presented on the screen. With the selected point spacing of 3 – 5 mm, a reliable identification of component boundaries is often possible with the intensity image alone (Fig. 5).

Jürgen giese technology mix in practice building documentation at bamberg cathedral
Jürgen giese technology mix in practice building documentation at bamberg cathedral

Fig. 5 Point cloud of a Faro Focus 3D displayed in Faro Scene. Average point spacing 5 mm. Assembly: N. Weather

However, if the differences in brightness between neighboring components are weak, the intensity values are often not sufficient for a reliable delimitation. However, the quality of the intensity values is influenced not only by the unalterable object properties, but also by currently existing software weaknesses: Compared to the representation of the intensity values in the proprietary scanner software, there is often still a significant loss of quality in the representation in the CAD system, which then also makes the evaluation based only on point clouds difficult or impossible (Fig. 6). As a way out of this problem, two variants are offered for further development:

1. The representation of point clouds in the CAD system is improved. 2. An interface is set up between the scanner software and the CAD system, with the help of which coordinates are tapped from the point cloud in the scanner software, which are then sent to the drawing tool active in the CAD system. Further software development will show which variant is the best to handle.

Jürgen giese technology mix in practice building documentation at bamberg cathedral
Jürgen giese technology mix in practice building documentation at bamberg cathedral

Fig. 6 Point cloud of a Faro Focus 3D displayed in Autodesk AutoCAD. Average point spacing 5 mm. Mounting: N. Weather

If the intensity values are not sufficient for a reliable identification and demarcation of components, one or more of the already existing photos can be oriented into the object coordinate system. The drawing process then takes place directly on an oriented photo, which has the point cloud as a reference surface for generating three-dimensional coordinates (Fig. 7). This process is also called monoplotting. With this method, it is also quite easy to extend the reference surface used into areas of scan shadows and thus to arrive at a complete evaluation. Scanning shadows are regularly present especially above ledges and cornices and cannot be avoided there. Photogrammetry is therefore basically required for a complete recording of components lying at a greater height, since only one camera can be lifted to the heights then required (Fig. 8).

Jürgen giese technology mix in practice building documentation at bamberg cathedral
Jürgen giese technology mix in practice building documentation at bamberg cathedral

Fig. Combined evaluation of point cloud. Oriented photo using Kubit PointCloud under Autodesk AutoCAD. Mounting: N. Weather

Jürgen giese technology mix in practice building documentation at bamberg cathedral
Jürgen giese technology mix in practice building documentation at bamberg cathedral

Fig. 8 Bamberg Cathedral, southwest tower. View of the top floor from the south. Photo from drone, flight altitude ca. 52 m above ground. Photo: Jürgen Giese

However, the orientation of photos, which in our case originate from cameras that cannot be fully calibrated in advance [8], can be a time-consuming procedure that can slow down the evaluation process considerably, taking 2-3 hours per photo. This is not so much due to the software used, but to a circular reasoning in the procedure itself: The oriented photo is always interesting when the intensity image of the point cloud is not sufficient for the evaluation, at the same time 10-15 – with inaccessible surfaces throughout natural – control points are needed for the orientation. If these control points are obtained from the point cloud, precisely defined points are searched for in an intensity image, precisely because only a few points can be precisely recognized in the intensity image. In practice, the orientation usually works despite this circular argument, but it costs the mentioned time effort to find the suitable points. The only alternative is to measure the natural control points on site with a total station, but this requires a similar amount of time and effort, including set-up, dismantling and stationing.

Jürgen giese technology mix in practice building documentation at bamberg cathedral
Jürgen giese technology mix in practice building documentation at bamberg cathedral

Fig. 9 Bamberg Cathedral, on-site verification and supplementary measurements with online tachymetry. Tachymeter Leica TS02 in combination with Kubit TachyCAD under Autodesk AutoCAD. The evaluation of point clouds. Photos taken in the field require approx. 60 percent of the total working time of the project.

All drawings created in the office require subsequent revision on site. Incorrectly interpreted findings are corrected and gaps in data acquisition are closed. The methods used for this purpose are online tacheometry and online surveying. The hand measurement used (Fig. 9 and 10).

Jürgen giese technology mix in practice building documentation at bamberg cathedral

Fig. 10 Bamberg Cathedral, coffin walls in the roof space. Verification and supplementary measurements on site by hand measurement in areas difficult to access. Photo: Jürgen Giese

Common misinterpretations in photos and point clouds concern, for example, the demarcation of stone surfaces from perfectly colored plasters or the overlooking of stonemason's marks (Fig. 11).

Jürgen giese technology mix in practice building documentation at bamberg cathedral

Fig. 11 Bamberg Cathedral, north wall of the central nave, detail from longitudinal section. Left Result of the evaluation in the office, right after correction and completion on site, changes are marked in red. Evaluation G. Kröck, revision on site A. Hager. Mounting:Jürgen Giese

The essential work on site, however, consists of closing data entry gaps. The scan shadows created by the three-dimensionality of the components can be reduced by a large number of setups and the addition of the photos described above, but they cannot be eliminated completely. In addition, there are areas in every point cloud, including the photos, that cannot be clearly interpreted and therefore are not evaluated. The complete recording of a building of the complexity of Bamberg Cathedral by terrestrial laser scanning alone proves to be an illusion at the latest now. Gaps and areas that cannot be clearly evaluated lurk z.B. At eaves, base zones, behind equipment parts, scaffolding positions and rainwater downpipes. Not to mention the numerous problem zones in the roof structure, where we completely dispensed with laser scanning right from the start and used only tacheometric and manual recording (Fig. 12).

The areas that can only be evaluated incompletely in the office often concern transitions between different groups of components, i.E. Zones of a building that are particularly important from a structural point of view. In order to achieve the goal of creating meaningful drawings that decipher the construction of a building, the precise observation and recording of precisely these zones is crucial. Only the consistent interrogation of the line drawing for a logical. The correct representation of the construction of all components reveals the deficits that can only be remedied by autopsy. Correction and completion of the drawings on site requires ca. 30 percent of the project's working time.

Jürgen giese technology mix in practice building documentation at bamberg cathedral
Jürgen giese technology mix in practice building documentation at bamberg cathedral

Fig. 12 Bamberg Cathedral, roof space, detail of longitudinal section. Measurement by tachymetry and hand measurement, representation still without line thicknesses and line types. Measurement Office for Building Research Peter Dresen. Office Geller – Bornschlögl. Assembly: Jürgen Giese

The combined use of terrestrial laser scanning, photogrammetry, tacheometry and hand measurement for the structural recording of such a large structure is a sensible and efficient way, provided that the interfaces between the methods are marked out according to their effectiveness . Despite all the technology, it must not be forgotten that the most important factor for the efficiency of a work process is the human being with his training and experience. Only after really solid training and extensive experience can one succeed in efficiently using the combinations of techniques outlined here. Through the experiences in teaching and in this project, it has become clear that the path to achieving this expertise is greatly facilitated when one and the same CAD system is used from the first evaluation of the point cloud of a tower helmet to the entry of the last manual measurement of a window fitting in the crypt. Only if all the usual construction and drawing tools are available at any time during the editing process, the drawings can be created efficiently.

For the future it is to be wished that the quality of the representation of point clouds increases and the handling with them becomes clearly more fluent, so that the identification of findings becomes more surely and the operational sequences with the tapping of coordinates approach still more to the tapping locally by means of tachymeter. With increasing computer power, it is expected that point cloud and photo evaluation can also take place on site, where many uncertainties in the finding approach can be easily answered and thus at least the step of checking the evaluation can be omitted.

Notes

[1] On the spectrum of construction documentation s. J. Giese in: K. Heine / K. Rheidt / F. Henze / A. Riedel (Ed.), From hand measurement to high-tech III (Mainz 2011) 122 f.

[2] More detailed on the Bamberg training objectives J. Giese in: K. Heine / K. Rheidt / F. Henze / A. Riedel (ed.), From hand measurement to high-tech III (Mainz 2011) 122-130.

[3] The project is supported by the professorships for building research and building history (S. Breitling) as well as restoration science in the preservation of historical buildings (R. Drewello) carried out in close cooperation with the State Building Authority Bamberg. Funding was made possible by generous support from the Upper Franconia Foundation. The author coordinates the project. Is responsible for quality assurance. The data collection on site and the evaluation is presented by N. Weather and A. Priests as scientific assistants, who are supported by numerous scientific assistants. Employed were and are H. Al Omar, V. Bauer, S. Bitrian, C. Eckstein, A. Hager, Ch. Henze, C. Kemna, L. Klahr, G. Kröck, Y. Kunisch, E. Micksch, J. Müller, T. Panke, J. Sharp, S. Screens and K. Vogt. We would like to take this opportunity to thank all the staff for their efforts. In addition, the Bamberg State Construction Office awards sub-projects to independent construction research offices, which are directly incorporated into the overall project..

[4] The scanner used is the Leica Scanstation 1.

[5] The used scanner is the Faro Focus 3D.

[6] A Leica TCRM 1201 total station in combination with forced centering was used to create the fixed point field. For the adjustment calculation the network adjustment module of the program Kubit TachyCAD was used. Cameras of the type Canon EOS 40D are used. A medium format camera Hasselblad H3D-39.

[8] S.O. Note. 7. On the current status of possible interfaces. Process combinations cf. Also the essays in K. Heine / K. Rheidt / F. Henze / A. Riedel (eds.), From hand measurement to high-tech III (Mainz 2011).

Like this post? Please share to your friends:
Leave a Reply

;-) :| :x :twisted: :smile: :shock: :sad: :roll: :razz: :oops: :o :mrgreen: :lol: :idea: :grin: :evil: :cry: :cool: :arrow: :???: :?: :!: