This is an attempt to recreate the EDITH glasses from Spider-Man Far From Home in a believable manner. And with the budgetary restraints of a broke high schooler (with generous parents).
Anyway, the main idea here is to implement facial recognition into effect and use display modules to display results in the eye lenses. This may not look polished, but its the closest we'll get with our current restraints.
This project is highly derivative of JLaser's video which can be found here. I wouldn't say it was inspired, I would call it direct copying:
https://youtu.be/Sj9ZQ20cgLA
For this project I'm using a couple of components that are listed below:
1) Display module:
https://www.sparkfun.com/products/15173
https://www.sparkfun.com/products/15123
Now that hardware is out of the way, let's talk software:
1) Python (I'm using Spyder IDE)
2) Arduino 1.8.5 (This version works best with required packages and modules)
Modules: i) face_recognition
ii) cv2
iii) NumPy
Arduino packages: i) HyperDisplay Library
ii) HyperDisplay SSD1309 Arduino Library
iii) HyperDisplay UG2856KLBAG01 Arduino Library
iv) U8g2
ii) HyperDisplay SSD1309 Arduino Library
iii) HyperDisplay UG2856KLBAG01 Arduino Library
iv) U8g2
So now that that's outta the way, let's talk progress. I'm not typing this while starting the project. The idea to blog the developments on this project only occurred to me a while into the project. As we stand now, the project has shown slow but comforting growth. It's probably slow because I procrastinate a lot.
The Python program requires an edit every time new faces are added to it. This problem could potentially be solved by the usage of automatic variable generators. I intend to integrate SQL into the program to help with the storage of data, and that might help with it.
Also, this process needs to be automated. As of right now, both programs need to be activated independently.
I also have plans on using Arduino WiFi boards to make the data transfer easier ( and to make the thing more portable).
In the future, I hope to switch to Raspberry Pi because it has OpenCV capabilities and may make the process easier.
That's it for now. Quail-Man flies.
Comments
Post a Comment