Start Here.

Current Progress

Before jumping into the how-to-start, it is important to have a good grasp of the current progress.

So far, a hardware prototype has been built and improved upon by past groups. This prototype has the camera and the proximity sensor already installed on it. The 3D printed attachment at the front of the prototype that a user might look inside has an LED strip installed on the walls to properly illuminate the eye. While it is fairly robust, this prototype, still lacks a few important things- a speaker/buzzer for audio feedback, weatherproofing, and mounts for the proximity sensor and the relay for the light. Furthermore, the the LED strip, when turned on, has some issues with glare which can be fairly significant for people with glasses. While these issues are quite small and wont affect operations too much, fixing them can really help improve the seamless automation envisioned by Professor Willow. Overall, significant progress has already been made on the hardware prototype and this device is very close to completion.

A majority of the software code for the project has also already been written. Due to the COVID outbreak, we, the Class of 2020 group were not able to test any of this old code. However, we have gone through all the files that Prof. Willow has kept track of in the Github repository and documented their theoretical use on this website. Finally, we have also marked which files are obsolete and can be safely ignored, along with their original intended use. Please note that none of these files have been delted for documentation purposes. This quick start guide should now correctly navigate you through the beginning steps of this project.

Quick Start Guide:

  1. Read through the background throughly and understand what the goal is. Look through this presentation to understand the project a little better. Make sure that you have also read through the project overview described on the home page and through the current progress written above. It is vital to understand these in order to succeed at this project
  2. Request Google Drive and Github access from professor Willow. Both of these have several documents that you might need to constantly refer to
  3. Look through the posters from 2018 and 2019 to try and get a big picture idea of what these groups have done. Here are a couple of key points from these two years:
    • The group in 2018 used a Haar Cascade classifier for Eye Detection
    • The group in 2019 used a Neural Network for Eye Detection
    • The group from 2019 completely scrapped the ideas from 2018 for a more robust and modern implementaion
    • We, the group of 2020 recommend sticking with the Neural Network implementation
  4. Now would be a great time to look at the code documentation. For your convenience, we have broken this up into two sections - the obsolete code from 2018 and the current code from 2019
    • Note: If the code documentation gets a little overwhelming, try ignoring any obsolete code and just looking at the most recent files
  5. Log-on to the main computer (the big one) that sits with Prof. Willow. Familiarize yourself with all the directory structures and locate all the files on the GitHub
  6. Try and get just the camera operating. The software for this should be on the computer already.
  7. Try and run the code from here. This might be hard to do, however once you have accomplished this, you should be ready to go on your own.
  8. It might also be a good idea to try an boot into the UDOO portable computer. The goal is to have this be the brains of the operations by just running a pre-trained Neural Network model.
  9. Finally, please make sure you follow any hyperlinks throughout this website. They are linked to datasheets, powerpoint presentations, and other helpful resources.
  10. Good luck! Make us proud!