CHI14 C10: Image Processing and Vision in Interaction and UX

posted in: CHI14, Conference | 0

This course is taught by J.R. Parker from University of Calgary

This course goes through OpenCV, Processing: More and more interfaces are being built that require visual input from the user. Certainly some of this can be accomplished using new hardware such as the Kinect, but computer vision provides a general solution absent special purpose devices other than a camera. After all, humans do the job without lasers or ultrasonic sensors. This course covers the essential vision and processing operations and outlines simple ways to make them work in practical situations. Easily accessible tools will be used that will allow the development of custom solutions.

This course is a good introductory course towards Vision & HCI. Personally, I do NOT recommend this course if you are already familiar with basic concept in Computer Vision & OpenCV.

Some of my notes:

  • 32-bit color = b + g<<8 + r<<16
    • b = c & 255
    • g = (c>>8) & 255
    • r = (c>>16)  & 255
  • Alpha = (pixel >> 24) & 255
    • pixel = b + g << 8 + r << 16 + alpha << 24
  • Threshold for OCR
    • Otsu’s algorithm (grey level histograms / iterative selection)
  • Difference between images for User Presence Detection
  • Color quad tree
  • Hue and Intensity Histograms
  • Re-quantization
    • Replace with a histogram of X colors
  • Vector templates for OCR from data glyphs
  • Hand gestures recognition
  • Text in the world
p1033-parker

Leave a Reply