Computer Vision Week 3 Low level vision Artifical

Posted on In
Symbols count in article: 774 Reading time ≈ 1 mins.

Task of this week

  • Edge and Feature detection
  • Requires Filtering
  • Requiers Convolution

Convolution

$$I’(i,j)=I*H=\sum_{k ,l}{I(i−k,j−l)H(k,l)}$$

  • To calculate values at each location in the filtered image:
  • You can imagine sliding the mask across the input image, filling in the values for the output (filtered) image as you go.
  • Alternatively, you can imagine the mask replicated at every pixel location in the output image, and the results generated in parallel (like the DoG filters in the retina).

Edge detection

  • Convolving with a differencing mask:

    • enhances edges
    • but also enhances noise

  • Convolving with a smoothing mask:

    • removes noise
    • but also blurs edges

  • There is a trade-off between edge detection and noise suppression

  • Edge detection requires:

    • a spatial scale, established using a smoothing operator(a Gaussian mask)
    • a differencing operator to find significant grey-level changes (at that spatial scale)