First Lecture, Computer Vision

January 19, 2010

I’m preparing my first lecture of the semester (to a class of 70 ?!), and planning to make the claim that computer vision algorithms really touch our lives much more than in the past.  So what are those algorithms?

  1. OCR/handwritten character recognition.
  2. Face detection in digital SLR cameras
  3. Face tracking in webcams
  4. smart phone apps like Google Goggles
  5. panoramic photo stitching
  6. iPhoto face annotation

What other recent algorithms are fully automatic and widely in use in consumer electronics?

Passive Vision.

January 10, 2010

Downtown St. Louis.

Welcome.  I plan to use this blog to ideas for both research and teaching.  At the moment, I’m looking forward to teaching Computer Vision to a class of 60 (!) in the Spring semester.  One of the techniques we will cover is PCA.  The above picture has two parts.  The left is one frame of a time-lapse from a camera set up in my lab.  The right image is a false color image, where the red, green, and blue channels are made from the first 3 principle components.

PCA has long been used for things such as face recognition, but also highlights structural components of a scene captured over time by a static camera.