Slides:



Advertisements
Similar presentations
Goal: a graph representation of the topology of a gray scale image. The graph represents the hierarchy of the lower and upper level sets of the gray level.
Advertisements

Computing Persistent Homology
ESTEC Photo Club How to improve Contrast & Brightness in your Photos Roberto Donadio TIA/TTG.
Introduction to Illustrator. With the release of Illustrator, Adobe has completed a tightly integrated trio of applications (Photoshop, InDesign and now.
Preliminaries Computing Betti Numbers Computing Persistent Homology.
Internet Vision - Lecture 3 Tamara Berg Sept 10. New Lecture Time Mondays 10:00am-12:30pm in 2311 Monday (9/15) we will have a general Computer Vision.
Topological Data Analysis
Multimedia for the Web: Creating Digital Excitement Multimedia Element -- Graphics.
Algorithmic Classification of Resonant Orbits Using Persistent Homology in Poincaré Sections Thomas Coffee.
Computer Vision Introduction to Image formats, reading and writing images, and image environments Image filtering.
Distinguishing Photographic Images and Photorealistic Computer Graphics Using Visual Vocabulary on Local Image Edges Rong Zhang,Rand-Ding Wang, and Tian-Tsong.
Lecture 1: Images and image filtering
Manifold Learning: ISOMAP Alan O'Connor April 29, 2008.
CS430 © 2006 Ray S. Babcock CS430 – Image Processing Image Representation.
Original image: 512 pixels by 512 pixels. Probe is the size of 1 pixel. Picture is sampled at every pixel ( samples taken)
Finding generators for H1.
Vector vs. Bitmap SciVis V
8.2 PICTURE IT Graphic Types and Their Uses. What Are Graphics?  Graphics are pictures/images  They could be images taken on a digital camera or scanned.
Sana Naghipour, Saba Naghipour Mentor: Phani Chavali Advisers: Ed Richter, Prof. Arye Nehorai.
Linear Algebra and Image Processing
Distinctive Image Features from Scale-Invariant Keypoints By David G. Lowe, University of British Columbia Presented by: Tim Havinga, Joël van Neerbos.
Abstract Some Examples The Eye tracker project is a research initiative to enable people, who are suffering from Amyotrophic Lateral Sclerosis (ALS), to.
Vector vs. Bitmap
Difference Between Raster and Vector Images Raster and vector are the two basic data structures for storing and manipulating images and graphics data on.
September 5, 2013Computer Vision Lecture 2: Digital Images 1 Computer Vision A simple two-stage model of computer vision: Image processing Scene analysis.
Topological Data Analysis
Computer Vision Lab. SNU Young Ki Baik Nonlinear Dimensionality Reduction Approach (ISOMAP, LLE)
1 Formation et Analyse d’Images Session 7 Daniela Hall 25 November 2004.
Segmentation: Region Based Methods Region-based methods –iterative methods based on region merging and/or splitting based on the degree of similarity of.
Persistent Homology in Topological Data Analysis Ben Fraser May 27, 2015.
Fields of Experts: A Framework for Learning Image Priors (Mon) Young Ki Baik, Computer Vision Lab.
Creating a simplicial complex Step 0.) Start by adding 0-dimensional vertices (0-simplices)
This work was partially supported by the Joint DMS/NIGMS Initiative to Support Research in the Area of Mathematical Biology (NSF ). Isabel K. Darcy.
Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert.
MATH:7450 (22M:305) Topics in Topology: Scientific and Engineering Applications of Algebraic Topology Sept 9, 2013: Create your own homology. Fall 2013.
A filtered complex is an increasing sequence of simplicial complexes: C0 C1 C2 …
A filtered complex is an increasing sequence of simplicial complexes: C 0 C 1 C 2 … UUU.
Guilford County SciVis V104.03
Lecture 1: Images and image filtering CS4670/5670: Intro to Computer Vision Noah Snavely Hybrid Images, Oliva et al.,
MATH:7450 (22M:305) Topics in Topology: Scientific and Engineering Applications of Algebraic Topology Sept 16, 2013: Persistent homology III Fall 2013.
Recombination:. Different recombinases have different topological mechanisms: Xer recombinase on psi. Unique product Uses topological filter to only perform.
Date of download: 6/26/2016 Copyright © 2016 SPIE. All rights reserved. Three different codebook sources have been used, the Lena image and the prior two.
Ancona - Municipality Hall The principal point PP is the intersection of the three heights of the triangle with vertices the three V.P. The V.P. procedure.
Recombination:. Different recombinases have different topological mechanisms: Xer recombinase on psi. Unique product Uses topological filter to only perform.
Sept 25, 2013: Applicable Triangulations.
From Natural Images to MRIs: Using TDA to Analyze Image Data
Application to Natural Image Statistics
Zigzag Persistent Homology Survey
Journal of Vision. 2003;3(5):3. doi: /3.5.3 Figure Legend:
We propose a method which can be used to reduce high dimensional data sets into simplicial complexes with far fewer points which can capture topological.
Oct 16, 2013: Zigzag Persistence and installing Dionysus part I.
Sept 23, 2013: Image data Application.
From: Optimum spatiotemporal receptive fields for vision in dim light
Application to Natural Image Statistics
plosone. org/article/info%3Adoi%2F %2Fjournal. pone
Defn: degree of si = deg si = time when si enters the filtration. .
CIS 601 Image Fundamentals

Clustering Via Persistent Homology
A filtered complex is an increasing sequence of simplicial complexes: C0 C1 C2 …

2.02C Frame-by-Frame Computer Animation Using PowerPoint
Principal Component Analysis (PCA)
Outline H. Murase, and S. K. Nayar, “Visual learning and recognition of 3-D objects from appearance,” International Journal of Computer Vision, vol. 14,
Lecture 5: Feature invariance
Lecture 5: Feature invariance
Lecture 10: The Future May 17, 2019
Betti numbers provide a signature of the underlying topology.
Number of white streaks for the cobalt-containing alloy (upper, black lines) and titanium clips (lower, white lines) scanned in high mode (H) (open symbols,
Presentation transcript:

Application to Natural Image Statistics With V. de Silva, T. Ishkanov, A. Zomorodian

An image taken by black and white digital camera can be viewed as a vector, with one coordinate for each pixel Each pixel has a “gray scale” value, can be thought of as a real number (in reality, takes one of 255 values) Typical camera uses tens of thousands of pixels, so images lie in a very high dimensional space, call it pixel space, P

Lee-Mumford-Pedersen [LMP] study only high contrast patches. Collection: 4.5 x 10 6 high contrast patches from a collection of images obtained by van Hateren and van der Schaaf

Lee-Mumford-Pedersen [LMP] study only high contrast patches. Collection: 4.5 x 10 6 high contrast patches from a collection of images obtained by van Hateren and van der Schaaf Choose how to model your data

Consult previous methods.

What to do if you are overwhelmed by the number of possible ways to model your data (or if you have no ideas): Do what the experts do. Borrow ideas. Use what others have done.

Carlsson et al used

The majority of high-contrast optical patches are concentrated around a 2-dimensional C 1 submanifold embedded in the 7-dimensional sphere.

0.) Start by adding 0-dimensional data points Persistent Homology: Create the Rips complex is a point in S 7

For each fixed  create Rips complex from the data 1.) Adding 1-dimensional edges (1-simplices) Add an edge between data points that are close is a point in S 7

For each fixed  create Rips complex from the data 2.) Add all possible simplices of dimensional > 1. is a point in S 7

For each fixed  create Rips complex from the data In reality used Witness complex (see later slides). 2.) Add all possible simplices of dimensional > 1. is a point in S 7

Probe the data

Can use function on data to probe the data

Large values of k: measuring density of large neighborhoods of x, Smaller values mean we are using smaller neighborhoods

Eurographics Symposium on Point-Based Graphics (2004) Topological estimation using witness complexes Vin de Silva and Gunnar Carlsson

Eurographics Symposium on Point-Based Graphics (2004) Topological estimation using witness complexes Vin de Silva and Gunnar Carlsson

From: From: Klein Bottle

M(100, 10) U Q where |Q| = 30 On the Local Behavior of Spaces of Natural Images, Gunnar Carlsson, Tigran Ishkhanov, Vin de Silva, Afra Zomorodian, International Journal of Computer Vision 2008, pp 1-12.

Combine your analysis with other tools