This week we shifted gears a bit while waiting for the RS-485 dummy implementation to be ready. The RS-485 Half-Duplex transmission is now fully operational between the Uno and Due and is ready to be translated for the Jetson/Due bus. While waiting for this to be done, other teammates focused on the assembly of the glove and headset, finally reading values in from the glove sensors, and fine-tuning the camera parameters of our webcam to mitigate motion blur issues.
The design for the camera mount on the helmet has been finalized, and an initial glove prototype is in the process of being assembled. Post-processing of acceleration data is in progress. After extensive research and discussions, we decided to change the acceleration into velocity data for them to get mapped with various parameters in a software synthesizer. We also acquired binary data from the force pressure sensors. These data will be used next week for establishing a connection between the Micro and the Due. Additionally, at the end of last week we noticed that motion blur was causing the glove tracking software to lose track of the fiducial markers if the camera or glove was moved too fast. We realized that this motion blur could be eliminated by raising the shutter speed of the webcam and balancing other exposure properties to compensate. Thus, Tomas spent most of the week figuring out how to adjust these parameters and fine tuning them for an improved tracking response.
As of now we are still slightly behind schedule, but we’ve made up a lot of our time from last week. Finally getting over the initial RS485 hurdle means that we can begin the full integration of our subsystems, and while we’re a bit behind schedule we are still confident that we have allocated enough slack time to meet our deadlines. Additionally now that all of our sensors are reporting data, all that needs to be done is to finish post-processing that data into usable figures. The CV portion of our system is fully operational, so most of the work on that front will be focused on improvement of our model.
This next week will see the beginning of the integration phase of our project. We hope to have RS485 communication working between the Jetson and the Due by the end of the week. We also hope to have some usable data from the glove sensors so that we can start integrating it as well in the following week. Finally, we will begin work on a more finalized design of the glove, attaching the Micro and fiducial markers.