Skip to the content.

Exercise 1

Part B: Get to know your data - start by reading the metadata and reports

With any sort of dataset, we need to assess the quality of the data. There are a number of considerations to take into account when assessing the 3D datasets you have at your disposal. The first step is to obtain any reports or metadata concerning the dataset. Metadata is information that describes other data – this may describe something as simple as the file format in which the data is saved, what equipment was used to record the data, or what other files are associated with the data. Understanding your data is essential as there are several characteristics that will dictate which software packages will be most useful in managing and processing the dataset for use.

Reading and Understanding the Metadata Report

Think and Respond: Summarise what you now know about the dataset after reading the metadata report. What are some of the challenges you can see in working with this dataset? Take note of what you think is most important and add it to your Design Document under the Resources: Available section.

Get to know your data – open and visually review samples

A step-by-step guide from this point onward in Exercise 1 can be found here.

Now that we know what type of data and what format the files are saved in, we need to identify how much of the site is represented in the record. First, we need to open the dataset in software well-suited to dealing with point clouds in the .las or .laz format. There are a number of paid software packages that will work with these files (Leica Cyclone, Autodesk ReCap, Esri ArcGIS), but there are a number of free and reliable open source software packages available. In this case, download and install the relevant version of CloudCompare.

Screencapture of the CloudCompare dialog box as it appears after importing a .las file.

The second dialog box that appears to translate the point cloud from a global coordinate system to a local coordinate system. A red rectangle highlights that the 'Preserve Global shift on Save' box should be unchecked.

We should also explore the other 3D datasets available in this archive to see if they will be useful in our reconstruction. Download MeshLab (or use CloudCompare to open the mesh, but note that the colour will not be applied). On Zenodo, search for Malthi SfM – this will come up with the Structure from Motion photogrammetry records produced in 2015 to create detailed 3D models of two wall segments at the site. Photogrammetry is a different method of 3D data capture which involves taking overlapping photographs of an object or structure, which software then uses to build the 3D geometry of the feature or object.

Think and Respond: Now that you have viewed the different point clouds and meshes available to you, assess how much of the data is needed for your reconstruction. What do you notice about what was captured in the scan data? How much of the site has been recorded? According to the report, everything within the fortified wall was recorded, though there may be related structures situated outside of this boundary. Will the trees and grass captured in the laser scans help or impede your reconstruction? It seems unlikely that an occupied settlement would have been covered in grass, so we can plan to remove these later. What level of detail will you require for your reconstruction – will the outline of buildings be sufficient to build from? While the snapshots of the full-resolution point cloud in the report show that the laser scans have recorded individual stones quite clearly, you need to decide if that level of detail is necessary for what you need to achieve. To meet this design brief, do you need the entire dataset, or only subsections of the recorded data (ie one house)? What size of dataset will be easy to work with (considering your computer’s hardware; if you find you are having issues with this, you may need to use one of the computers in UofG’s Digital Archaeology lab)? Do you need meshes or points? While point clouds appear quite attractive, the software we will be using and games engines [i.e. 3D modelling software which will be further discussed in later Exercises, and Unity and Unreal engines] are not generally designed to work with point clouds in this way. Summarise and incorporate your thoughts on these considerations in your Design Document template, particularly in the ‘Resources’ and ‘Constraints’ sections.

Back to Exercise 1 Part A Continue To Exercise 1 Part C