Poster
Posts
Week 4 - One Last Time
Objectives for today: Finish software with output logic Finish setting up hardware Make a poster Add more functionalities (optional) We had run into a problem right from the start as our PCB was printed out wrong. We hope that our PCB will arrive by today, but if worse comes to the worst we will have to finish it in the second group's lab schedule or make the output only single-digit. We tested the one-digit PCB and it works well. This would be our final product if we don't manage to get the second PCB in time. We are still working on improving the software. In the end, we had to be satisfied with our one-digit output.
Week 3 - I Ain't Worried
Objectives for today: Finish the basic structure of the software Put everything on PCB We started working on the code immediately and quite quickly finished the basic structure. Travis's method of using angles between finger landmarks was the most suitable to our system. We will also need to create output logic in the code to connect it to the hardware. The circuit design is also finished, we will solder after we are done with testing the final code output. The test circuit for 1-digit numbers was quickly connected to the Raspberry PI and worked wonderfully. We proceeded to build a two-digit test circuit right away. In the meantime, we are also debating about adding some functionalities to our code. The two-digit output test on the Raspberry PI caused us a lot of problems and we haven't managed to finish it this time. We will keep working on the software next week, and hopefully, it will take us little time as we would like to add some extra features an...
Week 2 - Try Everything
Objectives for today: Finish the hardware circuit design Continue working on the code We all came to the labs today. Our objectives for this session are quite difficult to measure whether we have completed it or not. We kept working on the code and hardware. We kept tabs on each other, sharing our ideas and tested a few methods for finger detection. The methods included using Mediapipe, contour detection with high contrast between hand and background. Using Mediapipe proved to be the simplest as it automatically assigns landmark to a hand. We are working with the landmark relationships. We also made significant progress with circuit design and tested it's number output using a testing code that counts from 0 to 9. Ben also designed a potential PCB for our circuit.