9/23/2023
Returned project to end state of Fall 2023.
All members now have 24/7 access to the project via Teamviewer.
Downloaded preloaded YOLO assets and weights for object detection.
Ensured that code could be edited in real-time along with the head Arduino for debugging and transmission purposes.
9/29/2023
Yolo premade weights were not good enough for our application. Using Roboflow, weights are in process of being crated manually via pictures.
Yolo V8 was installed on the Raspberry Pi.
Helped Joey to set up the USB Camera for the USB Camera versions of this code.
Designed a potential BREAD Slice design based off of the schematic acquired from Ian Huff from BREAD TTS.
10/7/2023
The unique weights in Roboflow are completed and transferred to the Raspberry Pi.
Spectrometer mount V1 has been successfully printed and works very consistantly.
Yolo V8 can now run on a local host via Docker and interpret photo feed via a confusion matrix based on model training.
Brett has completed a schematic for a potential Pi Vision slice that we will be printing soon and implementing on our Raspberry Pi.
10/28/2023
The spectrometer mount has been reprinted, and it is an significant improvement from the first print.
The BREAD Vision system can now run using YOLO with Color Detection at the same time.
The design for the Slice schematic has been updated, and the schematic will be given to Shane early this week for printing.
In preparation for the new physical system, a longer ribbon cable is to be purchased early this week.
11/5/2023
The PI Slice is now on order after quite a few adjustments to the design.
Spectrometer I2C connection in is development currently, waiting on new ribbon cable as current one is too short to use with mount.
The Roboflow model is being retrained with more stagnant positioning to better analyze imperfections.