Ptá se na věci, co jsou v Summary

Lecture 1

• Image
  • Continuous
  • Digital
  • How to derive from disc with Taylor series
• Cont → Disc: Taylor expansion
  • Approx the cont derivative using Disc formulas
  • Derive the forward/backward/central difference formula (01/21)
  • Precision of the approx (by the order - u O(h^2) se mění mocnina)
• Zbytek není tak důležitej

Lecture 2 - Linear Diffusion

• How the diffusion equation works (02/07)
• Analytic solution of the equation
• Gaussian filter → Convolution/FFT/Diffusion equation
• Time ~ Sigma parameter
• Can ask properties of Scale space
• Discretizing of the formula, how we solve the eq (cont→disc)
• Neumann boundary pochopit

Lecture 3 - Nonlinear Isotropic Diffusion

• What is diffusion, How it is defined (physical bg, 03/15), Fick’s law
• Diffusivity (g) - what it describes, flux (-g * grad(u))
• u_t = div(g * grad(u)); u(x, 0) = f(x) ⚠️ (initial condition)
• g is scalar now, a tensor in NAD
• Classification - differenced between diffusions, how to do NID with NAD, how to set the tensor D at that case, …
• Different diffusivity functions (3 co tam jsou, vědět, že můžou být další), Contrast parameter (what it does = how large diff you will have at a particular point), smaller lambda ~ smaller diffusivity