UAV and Photogrammetry

The advanced training conducted by the UCL team (30 Oct-1 Nov) illustrated the potential of using photogrammetry and UAV (drone) to rapidly record artifacts, buildings and sites within a holistic approach to heritage protection that moves from documentation to management. Carefully designed to complement the training offered by Durham in GIS and KCL in condition assessment, Dr Gai Jorayev worked closely with Drs Hiba Alkhalaf and Marco Nebbia to create a tiered level of complementary exercises. Prior to leaving the UK, the UCL team updated the training presentations and manuals, prepared several sample collections of images and other relevant data to be given to the participants. These were all shared with the participants using their USB flash drives.

Day 1

The training began with an introduction to higher complexity digital documentation suitable for publication, scholarly analysis, site presentation, outreach and site management plans. Through short and illustrated presentations, the UCL team showed the potential of photogrammetric recording as well as aerial photography. The examples included the final output that are ready to be integrated into GIS packages and interrogated further in the ways that were presented in previous days of the training by Durham University.

The afternoon was used to introduce small Unmanned Aerial Vehicle (UAV) to the group. The pros and cons of the very agile, inexpensive and easy-to-use DJI Mavic Pro UAV was demonstrated first in the classroom environment and then outside. The emphasis was on explaining safety features and the importance of controlling every aspect of flying and picture taking. Every participant had a chance to take-off, take some still images and videos, and then land.

Figure 1. Classroom session of introducing UAV

All of the participants demonstrated good understanding of the basic principles, although it was not possible to fly on a large archaeological site due to permit restrictions in Tunisia.

Figure 2. Outdoor UAV flying exercise.

The participants also had an opportunity to try out processing of UAV data from large site surveys, using the data collected at Iunca at the previous advanced training in February 2018.

Figure 3. Data from Yunca area processed by the participants.

Day 2

The first part of the day was dedicated to a practical documentation exercise at the site of Thyna using digital cameras and pre-prepared photogrammetric scales that had been built by the UCL team in London. The participants were divided into three groups and each group had a specific area of the site to model photogrammetrically. They had two outcomes in mind: 1) the outputs should be integrated into GIS and interrogated using the methods taught by Durham University team; 2) the outputs should have a practical value to be used in real-life condition assessment exercise to be undertaken as part of the training by KCL in coming days.

Figure 4. Participants working in Thyna.

Afternoon session in classroom showed that the participants achieved those two goals in all three groups. Processing photogrammetric data was undertaken in differing speeds because of the approaches to picture taking and computer power available, but all groups were able to demonstrate their results to the peers and discuss the challenges with them openly. Some of the models were of very high-quality and detail and can be used as element of long-term archive too.

Figure 5. Processing output example from Thyna data.

Day 3

This day was used to process photogrammetric data and to generate the outputs that are ready to be used as part of the condition assessment process. It provided a very good switchover the to the trainings of KCL in the coming day.

The day also was used to discuss ‘change over time’ analysis using photogrammetric techniques in order to monitor condition of artefacts and buildings. Gai Jorayev showed examples from elsewhere and also a quick test was carried out to show the potential using one of the pots within the premises of the training venue.

Figure 6. Change over time modelling.

UAV data from previous days was also used to create quick models of small areas and the value of aerial photography for 3D mapping was explained in detail.

Figure 7. Output of UAV flying from Day 1.

Dedicated UAV training

During the last two days of the training, four Tunisian participants were trained extensively in planning the data gathering with UAV’s, flying and processing the collected data. Due to difficulties of gaining permits, these training exercises primarily took place in a small area, but nevertheless, it was useful for the participants to gain more experience in this area. The UAV that was handed over to INP at the end of the training will be used by these trained specialists.

Figure 8. Participants presenting their processing results to the group.

Feedback during the sessions about the holistic approach to heritage management was very positive, and it is clear that combining GIS and photogrammetry (and UAV survey where possible) in order to use it in condition assessment exercises serves both as an excellent training course and as a feasible model for heritage management in Tunisia and Libya. It was also very encouraging to hear that certain participants found photogrammetric techniques very useful for their day-to-day work.