This is an old revision of the document!
The original idea
In order to engage users for longer term relationships it's important to be able to tell if you've met the user before, and recall past engagements.
- Need to be able to tell which user is which
- Replacement for rfid (which is too short range for unobtrusive use)
- Distinguish between a few people (<10)
- Use the camera and OpenCV's haar face finder
- Communicate over yarp
Also provide a test competency to try out reusable standard stuff for lirec (yarp, cross platform development, etc)
Initial version
- Using squared sum of differences (SSD) to compare stored images with any faces found in the image.
- Faces can be any size in the image, but are smoothly scaled into a standard resolution for comparison and storage
- Two modes of operation
Calibration Mode
Look for all faces in the image, compare with stored faces and create a new identity if the difference is over a given threshold. Do this each frame until switched into detect mode.
- Uses a maximum count of faces, in order to prevent storing too many false detections
Detection Mode
Look for all faces in the image, compare with stored faces and print the id number of the most similar stored face, if it's over a given threshold.
Demo Movie