Congratulations to Maximilian Heider for finishing his Bachelor Thesis!
After Maximilian Heider defended his bachelor thesis yesterday, we congratulate him on the successful completion of his bachelor thesis. The topic of his thesis was “Development of a framework for handheld kinematics analysis with affordable digital cameras”.
Maximilian’s task was to develop a more affordable version of Marius Oßwald’s framework with low-budget cameras (PlayStation 3 Eye cameras). The framework allows the recording of a subject’s hand performing a task with multiple cameras simutanously. With these recordings, it is possible to extract the hand kinematics with a neural network. Neural networks have shown in recent years that they can contribute in many research areas, and this is also true for markerless motion tracking.
A challenge of this bachelor’s thesis was to achieve high resolution of the video data while maintaining a high frame rate per second. The constraints were that the total cost of the setup should not have exceed 100€ and also be accessible to everyone. To understand the accuracy of the low-budget system he developed, he compared it to the already developed high-budget system in our lab. The accuracy was calculated with the mean-per-joint-position error for 400 test frames recorded with both frameworks.
In conclusion, Maximilian showed that the low-budget camera system provides very good results for motion analysis as long as the recorded motions are slow rather than fast.
We wish Maximilian all the best for his further studies and thank him for his contribution to our lab!