An Update on the Building Research

After taking part in two campaigns on Sai Island (see: The Architectural Survey Part 1 and Part 2 and the 3D-Laserscanning Campaign), I am now happy to announce that I have joined the team of AcrossBorders in the beginning of October, which enables me to fully focus on the architectural research of the New Kingdom town on Sai Island.

In the past months, my colleague from the Technical University of Vienna Robert Kalasek figured out a workflow for the post-processing of the huge amount of scan-data that we had acquired during our stay on Sai Island in February. This post-processing includes steps such as registering the single scans together and then cleaning the resulting 3D point cloud, i.e. removing any unwanted information and noises. After carrying out these steps and taking certain vital settings, such as the deviation, the range and the reflectance into consideration, a smooth data transfer into a further post-processing software was possible. In our case we are using the software PointCab, which enables us to create plans and sections directly from the 3D point cloud, which can be further worked out in AutoCAD.

Haus3 PointCab und Scan

3D point cloud of House 3 and plan of house 3 generated in PointCab

After Robert had already prepared some of the ground plans, my first task was to completely revise the ground plan that I had generated last year. The “old” plan was solely based on the sketches and measurements that I had carried out in my first year on Sai and therefore was lacking in accuracy in some parts. With the new laser scanning data, it was now possible to create an exact and – most importantly – a geo-referenced ground plan of the southern area of the Pharaonic town. At this point I would like to stress however, that my hand measurements and sketches were not in vain, since I believe that to really get to know a building and the structure you have to look really closely and carry out old-fashioned hand measurements to “get a feel” for it. Together with the 3D-data, these basics of documentation serve as the perfect combination for drawing analytical plans.

In a further step it is planned to integrate the newly excavated areas (SAV1 East and SAV1 West) as well as SAV1 North into this new plan. On the one hand, these areas were also scanned and geo-referenced, on the other hand they were exactly documented in the course of the excavation. Both of these results can easily be incorporated into the final map of the area. Also, work on detailed plans on the respective buildings in the form of accurate stone-by-stone (or rather brick-by-brick) plans has commenced and will be continued over the next few weeks.


3D scan of Building A (SAV 1 East)

Since a major part of my future work will focus on a reconstruction/visualization of the New Kingdom town, which shall have a strong scientific basis, I have started to look for comparative architecture in other settlements, e.g. Amarna and Elephantine in Egypt or Amara-West, Askut and Sesebi in Sudan. Different aspects of settlement architecture have to be taken in account, such as residential and administrative buildings, storage facilities, the fortifications and the entrance gate. The specific layout of the towns is being investigated as well, so that possible new conclusions for Sai can be made, especially taking into account that of now we still know little of the northern part of the town.

Going into more detail, other aspects I wish to elaborate on are the specific building types and also the building techniques used at the different houses.

H3_Überblick von Nord

House 3 as an example of a specific building type on Sai Island

In conclusion I can say that I am very much looking forward to the upcoming months and working with the architectural remains of Sai Island, which is certainly a very exciting field of work!

Looking back on the 3D-Laserscanning Campaign on Sai Island

Even though we have already been back in Vienna for some time, I would like to give some information on the 3D-laserscanning campaign, which was carried out on SAI Island from February 3-10, 2014. Robert Kalasek from the Technical University of Vienna was the main person in charge of the scanning process, providing his know-how and years of practical experience in the field of 3D-laser scanning. My part was to provide technical and scientific assistance, letting the knowledge acquired from last year’s architectural survey flow into the work process.

The main goal of the work was to obtain a complete geometric documentation of the remaining walls and floors of the southern part of the New Kingdom town on Sai Island. 3D-laserscanning seemed to offer the most comprehensive result which could be achieved in a relatively short time.

For the scanning an Image Laser Scanner Riegl VZ-1000 was used, a Nikon D800 camera with a 14 mm lens was mounted on the scanner in order to record the texture. During the scanning process a grid of three dimensional points is automatically measured in the surveyed area by the instrument. So-called 3D point clouds result from this process, including xyz-coordinates and an intensity value depending on the surveyed material.


Image Laser Scanner Riegl VZ-1000 in front of the ruins

The complete scanning of the remains of the Pharaonic town required 155 different scan positions, whereby the maximum distance of the measured points ranged between 200 and 400 m, according to the angle of incidence and the reflectivity of the material. The result of each scan is a point cloud in a local coordinate system. In a next step, the scans can be joined (registered) with the help of a multitude of reflector points, which were distributed throughout the ruins. Generally, at least five overlapping points are needed in order to put two scans together. These reflector points were also measured with a total station so that the registered scans can be placed into a geo-referenced net. With the help of the intensity information one can encode the three dimensional point clouds with different colors, which is often sufficient to differentiate the various objects.


Scan of part of the storage rooms of the Pharaonic town with different colors for the intensity of the textures

In addition to the standing remains of the Pharaonic town, the newly excavated trenches (SAV1 East and SAV1 West) as well as SAV1 North were also scanned and geo-referenced. In order to collect data for the topographic understanding of the surroundings, four long-range scans (range: 1.2 km) from elevated points were also undertaken. This could maybe help in understanding the general topography of the town and also to clarify questions such as the course of the city wall on the eastern side of the town.


Overview of the entire area of the Pharaonic town compiled from selected scans

Looking back on our days on Sai Island, I am happy to say that our campaign proved to be very successful and everything worked out as planned. Luckily, we had very few nimiti-days and the strong winds did not affect the stability of the scans as much as we had feared. The work-flow was very smooth and even though the work load was intense, especially for Robert, who worked on the data acquired during the day every evening until long into the night, we left Sudan with a lot of positive energy.

Next to the final registration of the scans, the post-processing in Vienna now entails the preparation of the data, e.g. reducing the enormous amount of data and refining the registration. The final output, a geometrically accurate three-dimensional model, can be used in the future to produce scaled plans and sections and a terrain model. It can also serve as a basis for further analyses and the creation of a reconstructed visualization of the Pharaonic town.