Exercise: Processing the Ezelke Dataset

This is a 3D model of the Ezelke (Dutch word for Donkey) sculpture created by Gertrud Januszewski for the city of Maastricht and its surroundings in 1976 to symbolise the green expansion of the city. The students who produced this 3D model as part of the course Computational Thinking for the Arts & Humanities (University College Maastricht) captured 158 images on a rainy and gloomy morning in September 2019. Some of these images are duplicates, dark or blurry, and not suitable to be used in the processing. Your task is to:

1. Download the dataset from here: https://shared.acdh.oeaw.ac.at/dh-moodle/4.2.3.Donkey/
2. Create a spreadsheet with four columns: photo number; keep; Discard; Reason
3. Go through the dataset and decide which images are worth keeping and which should be discarded
4. Note those that should be discarded in the spreadsheet always making sure that you justify the reason in the relevant column
5. Divide the good and bad images into two folders
6. Upload the good images to Agisoft Metashape and run an image quality estimate
7. Check if all your good images pass the 0.5 quality threshold. Those below 0.5 should be discarded.
8. Update your spreadsheet accordingly and move the low quality images (if any) to your 'bad' folder
9. Write a short blog post or a text for your reflective diary (c. 300 words) to summarise the quality of the dataset and explain what you might have done differently if you had the chance to record the sculpture yourself.

Ezelke by Maastricht 3D Collections on Sketchfab