Abstract

This master thesis focuses on the implementation and evaluation of an Augmented Reality guide in a museum. The main aim is to achieve a simple designed application with assistance elements that guides users through the museum. The system should expand the museum experience, be easy to understand, simple to use and joyful, to add to the normal visit. Selected statues were recognized with a camera to augment them. Other exhibits can be placed in the museum via ground detection to enable interaction with the exhibit. Both functionalities will disclose more information about the exhibit.
From the psychological and educational views, different research questions were defined to prove promising and increasing outcomes. Because of previous cooperation with the “Museum für Abgüsse Klassischer Bildwerke,” the application design will be redesigned with a focus on different user interface design principles. Another exhibit, the Column of Marcus Aurelius, will be integrated into the system.
Evaluation results of the application will be compared with a standard museum visit. The aim of the thesis is to figure out if an Augmented Reality guide can provide more information and motivation than regular attendance.

Project Description

The thesis will be carried out in the “Museum für Abgüsse Klassischer Bildwerke,” which contains multiple valuable plaster casts of famous ancient Greece and ancient Roman master-pieces. From the previous cooperation with the museum, a final integration of all previous works is needed to deliver a good impression to the visitors. This includes the AR projects about the Athena Parthenos statue from Madrid, the Crouching Aphrodite, the Nike of Samothrace, and at last, Laocoön and His Sons. Some reconstructions and contents of those projects were modified to fulfil the conditions with the museum. This application should be put online, so that each individual can choose whether to use the application when they visit.In addition, a new design is planned to introduce the user in an easier way. Different user interface principles will be covered to create a clear and structured guidance system. With regard to the new exhibition in the museum, some statues were selected that can be recognized with the Augmented Reality camera. In addition,the Column of Marcus Aurelius with the corresponding historical representation will be integrated into the final application. This exhibit can be placed with the help of ground detection. Surfaces will be recognized in the algorithm and marked to notify the user the areas that are detected as ground floor. When clicking on the area, the column will be deposit in front of the user. The two methods, the tracking process as well as the ground detection process, will provide multiple interactions and information about the target exhibit.

The application should be a self-explained museum guide to the user to help collecting important information of the exhibition and famous cast. In addition, the application should act as a good explained guide to the visitor thought the museum. Other challenges were the combination of previous projects into one application with all provide functionalities, the focus on a clear UI design, and the integration of a placeable exhibit, the Column of Marcus Aurelius. The involvement of the column should also center the defined research questions to achieve the best possible results. All learned UI decisions from Shneiderman and the official Google design will be integrated and shown in the actual developed application.

The column of Marcus Aurelius stands in Piazza Colonna, Rome/Italy. It consists of 27 or 28 blocks or Carrara marble, each is 3.7 m. The column’s high base measures around 6 m high, and the column’s shaft is around 30m. The column of Marcus Aurelius is a Doric column with a spiral relief and is a monument that tells the story of the Roman emperor Marcus Aurelius. In total, 116 scenes display the brutal truth of the Danubian or Marcomannic wars. Due to the pollution situation, acid rain, and other impacts, the column has become more and more damaged. For conservation arrangements of the column, the exhibit can be retained at its current form and reconstructed with the help of casts.

The main intuition behind this integration was to allow the visitor visualize the exhibit. Due to the size of the Column, it was not possible to place the complete exhibit in the museum,but this aspect should be fixed with the help of Augmented Reality. Ground detection tools allowed the user to first track the environment to capture possible flat surfaces and then place it in front of the visitor. Due to the fascinating size of the column, the exhibit should be explored by moving the target device to visualize the proportion. Once the Column of Marcus Aurelius was placed, the user could inspect and interact with it.

Results

The Between-subject study design was used: users are divided into several groups, each working with one alternative. In this case, two different groups are needed to compare the results and figure out which method achieves a better learning outcome. Group A contain the normal visitors as the control group. These participants will get some information sheets because the museum does not place several information panels around the exhibits.
Group B will contain participants who will receive the application guide at the beginning.

Unfortunately, due the coronavirus COVID-19 pandemic, we had to abort the user study because the museum was closed and it was not possible and allowed to continue the test phase. Hence, the given results are not complete. The amount of people in Group A were completed but more participants could improve the results. Especially, Group B is incomplete with a total of 7 participants.

To investigate the correlation between the groups and the gained exhibit knowledge score,the null hypothesis will be defined as “it exists a context between the group and the score”. Due to the null hypothesis, a relation is proven between both, if the given significant value(p-value)or alpha-value is 0.05 or lower; otherwise, the correlation is statistic accidental. The given error should be between 5% and 1% to accept the assumption or under 1%, which proves the statistically significant. Each of the following test methods contains a resulting p-value which predicts how precisely the statement is statistically significant.

The Kruskal-Wallis test between both groups reached a p-value of 0.01282.
The bootstrapped Kruskal-Wallis p-value of 0.01154 predicts a promising outcome.

Three significant values were scored due the quantile regression test: the 25th Percentile with p = 0.0048, the 50th Percentile with p = 0.0070 and the 75th Percentile with p = 0.0292.

Conclusion

Several challenges, such as the redesign in regard of multiple UI principles, the combination of all previous works in one project, and the integration of the placeable exhibit in a ground detection environment were mastered.

“I really like the intuition of the virtual guide. It is a really interesting method to gather information. I makes more fun to use the app than just read all given information on a plate.” - Anonymous Participant 5

In regard to the defined research questions and the shown TPACK model, this thesis was able to prove all statements with the given results,and p-values of p < 0.02 were found in all given evaluation methods, critics, and positive feedback. Due to the coronavirus COVID-19 pandemic and the corresponding fact that the number of participants was too small, a clear deduction cannot be made, but this approach is a good foundation. Future expanded evaluations can be made to improve the result to clarify the prediction.

[ Thesis PDF ]