example g x y e x2 y2 2 2 x 1

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: smoothing filter first Enhances edge areas Stage 3: Detec-on }༇  Produces a binary output, with edges marked by 1, other areas as zero Smoothing Masks 1 g ( x, y ) = ce 1 1 1 1 1 5 21 35 21 5 21 94 155 94 21 35 155 255 155 35 21 94 155 94 21 5 21 35 21 5 1/9 x }༇  − 1 1 Mean Filter }༇  Gaussian Filter 1 ( x2 + y2 ) 2σ 2 Filtering Gaussian Filter }༇  Pixel weight is inversely propor-onal to distance from origin }༇  g ( x, y ) = e }༇  ( x2 + y2 ) − 2σ 2 Value of σ has to be specified Gaussian Filtering }༇  Example g ( x, y ) = e }༇  − ( x2 + y2 ) 2σ 2 x = - 1, y = - 1, σ = 1 g (−1,−1) = e ( −1) 2 + ( −1) 2 − 2 0.367 ≈ 0.367 Gaussian Filtering }༇  Implementa-on problem }༇  }༇  Float mul-plica-ons are slow Solu-on }༇  }༇  Mul-ply mask with 255, round to nearest integer Scale answer by sum of all weights Gaussian Filtering 94 155 94 1 155 255 155 94 155 94 5 21 35 21 5 21 94 155 94 21 35 155 255 155 35 21 94 155 94 21 5 21 35 21 5 ≈ 2 1 2 4 2 1 2 1 Round(g/70) Gaussian Filtering Continuous Gaussian Function }༇  1- D Gaussian x2 −2 2σ 1 g ( x) = e 2π σ }༇  Non zero from - ∞ to ∞ }༇  Width is controlled by σ }༇  Symmetric }༇  Area under curve = 1 ∞ ∫ −∞ x2 −2 1 e 2σ = 1 2π σ Continuous Gaussian Function 2- D Gaussian Function g ( x, y ) = 1 2πσ e ( x2 + y2 ) − 2σ 2 Gaussian Filter ⊗ = Derivative Masks From defini-on of deriva-ves -1 1 -1 1 Prewi[ Operator -1 0 1 -1 -1 -1 -1 0 1 0 0 0 -1 0 1 1 1 1 }༇  Sobel Operator -1 1 -1 -2 -1 -2 0 2 0 0 0 -1 }༇  0 0 1 1 2 1 Robert’s Operator 1 0 0 1 0 -1 -1 0 Properties of Masks }༇  Filtering Masks }༇  }༇  }༇  }༇  }༇  All values are +ve Sum to 1 Output on smooth region is unchanged Blurs areas of high contrast Larger mask - > more smoothing }༇  Deriva-ve Masks }༇  }༇  }༇  }༇  }༇  opposite signs Sum to zero Output on smooth region is zero Gives high output in areas of high contrast Larger mask - > more edges detected Application of 2- D Masks If fx is deriva-ve in x- direc-on, fy is deriva-ve...
View Full Document

This note was uploaded on 03/02/2014 for the course CS 436 taught by Professor Sohaibkhan during the Winter '13 term at Lahore University of Management Sciences.

Ask a homework question - tutors are online