Low pass hl 19052009 dr pierre elbischger

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 5.2009 + - + + •(Dr. Pierre Elbischger - MIP1/ISAP'SS09 A-1) 7 High-boost filtering (2) For the low-pass filter, the Gaussian filter is often used. More robust to noise compared to the sharpening using the Laplacian Tunable parameters σ (spatial spread) and A (sharpening intensity) A=2 the expression is identical to unsharp masking (a=1, see previous slides) In general an arbitrary high-pass filter can be used. If we use the Laplacian for the high-pass filter, then for: A=1 the expression is identical to the (negative) Laplacian of the image A=2 the expression becomes identical to the sharpening using a Laplacian (see previous slides) 19.05.2009 Dr. Pierre Elbischger - MIP1/ISAP'SS09 8 Probability - Classical definition (Laplace 1812) A two-dimensional representation of the outcomes of two dice, and the subspaces associated with the events corresponding to the sum of the dice being greater than 8, or less than or equal to 8. 19.05.2009 Dr. Pierre Elbischger - MIP1/ISAP'SS09 9 Empirical probability (VonMises 1936) The experiment is repeatable forP ( A) := lim n event occurs nA times. n times whereby the n The probability is defined by the relative frequency nA/n. Due to the fact that n is always finite, P(A) can only be an estimate. Law of large numbers → For an increasing number of experiments, the probability approaches a limit. A n →∞ Example: Relative frequency (throw a dice n times) 19.05.2009 Dr. Pierre Elbischger - MIP1/ISAP'SS09 10 Distribution function FX (α ) : P( x ≤ α ) = Example: Dice 19.05.2009 Properties: d X (+∞) FX (−∞) F = 0 an= 1 FX (α 2 ) − FX (α1 ) P (α1 < x ≤ α 2 ) ≥ 0 = P(= α ) 0, if FX is continous. x= Dr. Pierre Elbischger - MIP1/ISAP'SS09 11 Probability density function The first derivative of the distribution function FX(α) is defined as the probability density function (PDF) fX(α) of the random variable x. dFX (α ) f X (α ) := dα Properties: f X (α ) ≥ 0 = FX (α ) α ∫ ⇐ monotony of the distribution function f X (γ ) ⋅ d γ in particular +∞ = FX (+∞) −∞ ∫ = f X (γ ) ...
View Full Document

This note was uploaded on 07/09/2009 for the course MEDIT 1 taught by Professor Pierreelschbinger during the Spring '09 term at Carinthia University of Applied Sciences.

Ask a homework question - tutors are online