Updates

Our team decided to work on a Smart Mirror for our capstone project. A smart mirror is a mirror with an LCD display behind it than can project useful information to the user, such as the weather, time, bus routes, or even a to-do list. We are inspired by the numerous possiblities this platform can support, which gives us a lot of freedom for creativity in this project. Our focus will be on implementing facial recognition and a mobile application that is integrated with our mirror.

Status: Waiting on hardware to start development of Smart Mirror features. Currently doing research on how to implement our desired features.

Plan: Meeting set for Tuesday 5 PM to meet with a Computer Vision expert for advice, development can start once we receive the hardware


Development

Here are the features we are currently researching for our mirror:

Facial Recognition

To make the mirror even ‘smarter’, we are going to incorporate a computer vision element into the design. By using a camera for facial recognition, we can identify the specific person standing in front of the mirror and load their own custom profile.

Currently we are looking into the OpenCV framework to help us obtain accurate facial recognition. OpenCV is an Open Source Computer Vision library that is well supported and documented, which makes it perfect for our needs. Soon, we are meeting with a referral to get some advice on this computer vision component of our design. Hopefully, we can get also some advice on a camera, so we can order that as soon as possible.

Mobile Application

We also want to integrate a phone application that can be used in conjunction with the mirror. The phone application could be used for all kinds of user interaction, such as setting up the layout of the mirror. We are also looking into more advanced features that the phone application could support, such as dual display or swiping gestures to push the phone contents to the mirror.

Gesture Control

We are considering some kind of gesture control to navigate the mirror. Our mirror is not a touch screen, and frankly we wouldn’t want it to be since it would just get smudged up by finger prints. Using the camera to detect simple gesture movements would be ideal, but we are also looking into using a standalone gesture interpreter such as the Hover.

Smart Mirror ON/OFF Protocol

  1. Smart Mirror is currently in OFF state, camera waits for movements
  2. Smart Mirror detects movement and turns ON, waits to recognize a face
  3. Face is recognized, and if a matching profile was saved, it is loaded; otherwise, a default profile is loaded
  4. User may now view information displayed on the mirror
  5. Face leaves view of the camera, there will be a buffer time of ~30 seconds before the mirror automatically goes back to its OFF state

Next steps

This week we are ordering a camera so we can begin implementing the computer vision features for our mirror. The other parts are not as time critical to our design, but we will also order the rest of the materials as soon as possible.

Here is a breakdown of what we are individually researching this week:

  • Brandon: Web application
  • Brad: Facial recognition
  • Ryan: Facial recognition
  • Kurtis: Gesture control / Facial recognition