Filters– Chapter 6. Filter Difference between a Filter and a Point Operation is that a Filter utilizes a neighborhood of pixels from the input image to.

Slides:



Advertisements
Similar presentations
3-D Computer Vision CSc83020 / Ioannis Stamos  Revisit filtering (Gaussian and Median)  Introduction to edge detection 3-D Computater Vision CSc
Advertisements

Linear Filtering – Part I Selim Aksoy Department of Computer Engineering Bilkent University
Spatial Filtering (Chapter 3)
Topic 6 - Image Filtering - I DIGITAL IMAGE PROCESSING Course 3624 Department of Physics and Astronomy Professor Bob Warwick.
CS & CS Multimedia Processing Lecture 2. Intensity Transformation and Spatial Filtering Spring 2009.
Image Filtering. Outline Outline Concept of image filter  Focus on spatial image filter Various types of image filter  Smoothing, noise reductions 
1Ellen L. Walker ImageJ Java image processing tool from NIH Reads / writes a large variety of images Many image processing operations.
5/24/2015© 2009 Raymond P. Jefferis III Lect Geographic Image Processing Data Transformation and Filtering Noise in data Arrays Digital filtering.
Lecture 4 Linear Filters and Convolution
Image Filtering CS485/685 Computer Vision Prof. George Bebis.
1 Image filtering Hybrid Images, Oliva et al.,
Image processing. Image operations Operations on an image –Linear filtering –Non-linear filtering –Transformations –Noise removal –Segmentation.
7. Neighbourhood operations A single pixel considered in isolation conveys information on the intensity and colour at a single location in an image, but.
1 Image Filtering Readings: Ch 5: 5.4, 5.5, 5.6,5.7.3, 5.8 (This lecture does not follow the book.) Images by Pawan SinhaPawan Sinha formal terminology.
CS443: Digital Imaging and Multimedia Filters Spring 2008 Ahmed Elgammal Dept. of Computer Science Rutgers University Spring 2008 Ahmed Elgammal Dept.
1 Image filtering Images by Pawan SinhaPawan Sinha.
Image Analysis Preprocessing Arithmetic and Logic Operations Spatial Filters Image Quantization.
Lecture 2. Intensity Transformation and Spatial Filtering
Linear Filtering About modifying pixels based on neighborhood. Local methods simplest. Linear means linear combination of neighbors. Linear methods simplest.
CSCE 441: Computer Graphics Image Filtering Jinxiang Chai.
CS 376b Introduction to Computer Vision 02 / 26 / 2008 Instructor: Michael Eckmann.
Presentation Image Filters
CS559: Computer Graphics Lecture 3: Digital Image Representation Li Zhang Spring 2008.
Spatial Filtering: Basics
Digital Image Processing Image Enhancement Part IV.
Basis beeldverwerking (8D040) dr. Andrea Fuster Prof.dr. Bart ter Haar Romeny dr. Anna Vilanova Prof.dr.ir. Marcel Breeuwer Convolution.
Filtering and Enhancing Images. Major operations 1. Matching an image neighborhood with a pattern or mask 2. Convolution (FIR filtering)
Digital Image Processing Lecture 5: Neighborhood Processing: Spatial Filtering Prof. Charlene Tsai.
Lecture 03 Area Based Image Processing Lecture 03 Area Based Image Processing Mata kuliah: T Computer Vision Tahun: 2010.
Image Processing Edge detection Filtering: Noise suppresion.
Image processing Fourth lecture Image Restoration Image Restoration: Image restoration methods are used to improve the appearance of an image.
Convolution and Filtering
Chapter 5: Neighborhood Processing
Linear filtering. Motivation: Noise reduction Given a camera and a still scene, how can you reduce noise? Take lots of images and average them! What’s.
Machine Vision ENT 273 Image Filters Hema C.R. Lecture 5.
Spatial Filtering.
Image Subtraction Mask mode radiography h(x,y) is the mask.
CSC508 Convolution Operators. CSC508 Convolution Arguably the most fundamental operation of computer vision It’s a neighborhood operator –Similar to the.
Digital Image Processing Lecture 16: Segmentation: Detection of Discontinuities Prof. Charlene Tsai.
Intelligent Vision Systems ENT 496 Image Filtering and Enhancement Hema C.R. Lecture 4.
Digital Image Processing Lecture 5: Neighborhood Processing: Spatial Filtering March 9, 2004 Prof. Charlene Tsai.
Image enhancement Last update Heejune Ahn, SeoulTech.
Linear filtering. Motivation: Image denoising How can we reduce noise in a photograph?
Digital Filters. What are they?  Local operation (neighborhood operation in GIS terminology) by mask, window, or kernel (different words for the same.
Instructor: Mircea Nicolescu Lecture 5 CS 485 / 685 Computer Vision.
Filtering (II) Dr. Chang Shu COMP 4900C Winter 2008.
CSE 185 Introduction to Computer Vision Image Filtering: Spatial Domain.
Sliding Window Filters Longin Jan Latecki October 9, 2002.
Spatial Filtering (Chapter 3) CS474/674 - Prof. Bebis.
Image Subtraction Mask mode radiography h(x,y) is the mask.
Linear Filters and Edges Chapters 7 and 8
Linear Filters and Edges Chapters 7 and 8
Basis beeldverwerking (8D040) dr. Andrea Fuster dr. Anna Vilanova Prof
Linear Filters and Edges Chapters 7 and 8
ECE 692 – Advanced Topics in Computer Vision
A Gentle Introduction to Bilateral Filtering and its Applications
Filtering – Part I Gokberk Cinbis Department of Computer Engineering
Digital Image Processing
Image filtering Images by Pawan Sinha.
9th Lecture - Image Filters
Digital Image Processing
Image filtering Images by Pawan Sinha.
Spatial operations and transformations
Digital Image Processing Week IV
Lecture 2: Image filtering
Image Filtering Readings: Ch 5: 5. 4, 5. 5, 5. 6, , 5
Lecture 7 Spatial filtering.
Digital Filters.
Spatial operations and transformations
Presentation transcript:

Filters– Chapter 6

Filter Difference between a Filter and a Point Operation is that a Filter utilizes a neighborhood of pixels from the input image to create a single pixel on the output image e.g. It’s known as a “multiply and accumulate” operation

Filters Parameters include –The size of the filter –The shape of the filter –The filter coefficients (weights) In the previous example the –The size was 3x3 –The shape was rectangular (square) –The coefficients were 1/9 (factored out for efficiency reasons)

Filters In matrix form Coefficients can be real numbers, positive, negative, and/or zero

Filters Every filter has a designated origin –The dimensions are typically odd numbers and the origin is the center coefficient –But this doesn’t have to be the case, dependant on what you are trying to accomplish

Filter u v 0 0 I u v I’ multiply add

Filter The multiply/accumulate operation is formally called convolution The ∞ is not a problem – just assume 0 valued coefficients The filter matrix is called the convolution kernel

Convolution Commutative – can swap the roles of the kernel and image Linear – distributive over addition (of images) and multiplication (by a scalar) Associative – order of operations is irrelevant Separable – some 2D kernels can be expressed as 2 1D kernels for implementation efficiency –The Gaussian is the most useful of the separable filters

Filter

ImageJ filters Process->Filters->Convolve…

ImageJ Filters Coefficients can be positive or negative, integer or real Normalize divides coefficients by the sum of the coefficients –It does something if the sum is 0, not sure what Make sure all rows are of the same length (don’t add an extra after the last row Filters can be saved as text files Filters can be read in/displayed as images or surface plots –File->Import->Text Image –Analyze->Surface Plot… for surface plot display –Filters with negative coefficients will be shifted (brightened) for display purposes (original values are retained in the image)

ImageJ Filters

Filters Can be arbitrary size and shape –Small filters are fast to run but provide only spatially local support Why might this be a problem? –May be dominated by noise –Large filters are slow to run but provide a broader spatial support Why might this be better? –Noise is averaged out over a larger region (statistics of larger samples) Dimensions are typically odd numbers to keep the origin centered Non-rectangular filters (e.g. circular) are embedded in a rectangle

Filters Image edges filter Image Border

Filters Ignore the edges Only apply the parts of the filter that land on image pixels Reflect the edges –This reduces the introduction of artifacts on the edges but is costly in terms of memory and time

Filters Infinitely many filters available, most are of no interest –Those of interest include Smoothing (blurring) –Average – all coefficients the same –Gaussian – coefficients create a Gaussian surface Gradient (difference) –Edge detection –Edge sharpening

Non-linear filters Linear filters treat all pixels the same –Smoothing out noise also smoothes out edges Nonlinear filters can avoid this situation –Min filter – replace the target pixel with the smallest value within the kernel neighborhood –Max filter – replace the target pixel with the largest value within the kernel neighborhood –Median filter – replace the target pixel with the central value of the sorted kernel neighborhood pixels (can be weighted) –Specialty filters – edge detectors, corner detectors, etc. (to be discussed later)