1 Information Extraction Principles for Hyperspectral Data David Landgrebe Professor of Electrical & Computer Engineering Purdue University

Slides:



Advertisements
Similar presentations
Applications of one-class classification
Advertisements

Component Analysis (Review)
Notes Sample vs distribution “m” vs “µ” and “s” vs “σ” Bias/Variance Bias: Measures how much the learnt model is wrong disregarding noise Variance: Measures.
A Graphical Operator Framework for Signature Detection in Hyperspectral Imagery David Messinger, Ph.D. Digital Imaging and Remote Sensing Laboratory Chester.
Major Operations of Digital Image Processing (DIP) Image Quality Assessment Radiometric Correction Geometric Correction Image Classification Introduction.
Multiple Criteria for Evaluating Land Cover Classification Algorithms Summary of a paper by R.S. DeFries and Jonathan Cheung-Wai Chan April, 2000 Remote.
Chapter 2: Pattern Recognition
Image Classification.
Classification and application in Remote Sensing.
Multispectral Remote Sensing Multispec program from PurdueMultispec On Information Extraction Principles for Hyperspectral Data, David Landgrebe, School.
Basics: Notation: Sum:. PARAMETERS MEAN: Sample Variance: Standard Deviation: * the statistical average * the central tendency * the spread of the values.
Lecture 14: Classification Thursday 18 February 2010 Reading: Ch – 7.19 Last lecture: Spectral Mixture Analysis.
Image Classification To automatically categorize all pixels in an image into land cover classes or themes.
Pattern Recognition. Introduction. Definitions.. Recognition process. Recognition process relates input signal to the stored concepts about the object.
Lecture 14: Classification Thursday 19 February Reading: “Estimating Sub-pixel Surface Roughness Using Remotely Sensed Stereoscopic Data” pdf preprint.
Aerial photography and satellite imagery as data input GEOG 4103, Feb 20th Adina Racoviteanu.
Chapter 12 Spatial Sharpening of Spectral Image Data.
EE513 Audio Signals and Systems Statistical Pattern Classification Kevin D. Donohue Electrical and Computer Engineering University of Kentucky.
Methods in Medical Image Analysis Statistics of Pattern Recognition: Classification and Clustering Some content provided by Milos Hauskrecht, University.
Remote Sensing Hyperspectral Remote Sensing. 1. Hyperspectral Remote Sensing ► Collects image data in many narrow contiguous spectral bands through the.
Machine Learning1 Machine Learning: Summary Greg Grudic CSCI-4830.
Environmental Remote Sensing Lecture 5: Image Classification " Purpose: – categorising data – data abstraction / simplification – data interpretation –
Speech Recognition Pattern Classification. 22 September 2015Veton Këpuska2 Pattern Classification  Introduction  Parametric classifiers  Semi-parametric.
1 Urban Growth Simulation A Case Study of Indianapolis Sharaf Alkheder & Jie Shan School of Civil Engineering Purdue University March 10, 2005.
A New Subspace Approach for Supervised Hyperspectral Image Classification Jun Li 1,2, José M. Bioucas-Dias 2 and Antonio Plaza 1 1 Hyperspectral Computing.
Seto, K.C., Woodcock, C.E., Song, C. Huang, X., Lu, J. and Kaufmann, R.K. (2002). Monitoring Land-Use change in the Pearl River Delta using Landsat TM.
Intelligent Vision Systems ENT 496 Object Shape Identification and Representation Hema C.R. Lecture 7.
Image Classification 영상분류
Overview of Supervised Learning Overview of Supervised Learning2 Outline Linear Regression and Nearest Neighbors method Statistical Decision.
Remote Sensing Supervised Image Classification. Supervised Image Classification ► An image classification procedure that requires interaction with the.
Jun-Won Suh Intelligent Electronic Systems Human and Systems Engineering Department of Electrical and Computer Engineering Speaker Verification System.
1 Particle Swarm Optimization-based Dimensionality Reduction for Hyperspectral Image Classification He Yang, Jenny Q. Du Department of Electrical and Computer.
1 E. Fatemizadeh Statistical Pattern Recognition.
Pattern Recognition April 19, 2007 Suggested Reading: Horn Chapter 14.
Chapter 8 Remote Sensing & GIS Integration. Basics EM spectrum: fig p. 268 reflected emitted detection film sensor atmospheric attenuation.
Map of the Great Divide Basin, Wyoming, created using a neural network and used to find likely fossil beds See:
Digital Image Processing
Hyperspectral remote sensing
Chapter 20 Classification and Estimation Classification – Feature selection Good feature have four characteristics: –Discrimination. Features.
Remote Sensing Unsupervised Image Classification.
Introduction to Machine Learning Multivariate Methods 姓名 : 李政軒.
Classification Course web page: vision.cis.udel.edu/~cv May 14, 2003  Lecture 34.
Sub pixelclassification
Computer Vision Lecture 7 Classifiers. Computer Vision, Lecture 6 Oleh Tretiak © 2005Slide 1 This Lecture Bayesian decision theory (22.1, 22.2) –General.
Integrating LiDAR Intensity and Elevation Data for Terrain Characterization in a Forested Area Cheng Wang and Nancy F. Glenn IEEE GEOSCIENCE AND REMOTE.
Part 3: Estimation of Parameters. Estimation of Parameters Most of the time, we have random samples but not the densities given. If the parametric form.
1 C.A.L. Bailer-Jones. Machine Learning. Data exploration and dimensionality reduction Machine learning, pattern recognition and statistical data modelling.
The MATLAB Hyperspectral Image Analysis Toolbox
Mapping of Coastal Wetlands via Hyperspectral AVIRIS Data
Nonparametric Density Estimation – k-nearest neighbor (kNN) 02/20/17
Hyperspectral Sensing – Imaging Spectroscopy
IMAGE PROCESSING RECOGNITION AND CLASSIFICATION
LECTURE 09: BAYESIAN ESTIMATION (Cont.)
LECTURE 10: DISCRIMINANT ANALYSIS
CH 5: Multivariate Methods
Map of the Great Divide Basin, Wyoming, created using a neural network and used to find likely fossil beds See:
University College London (UCL), UK
Outline Peter N. Belhumeur, Joao P. Hespanha, and David J. Kriegman, “Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection,”
Hyperspectral Image preprocessing
Classification Discriminant Analysis
Face Recognition and Detection Using Eigenfaces
REMOTE SENSING Multispectral Image Classification
REMOTE SENSING Multispectral Image Classification
Image Information Extraction
University College London (UCL), UK
Generally Discriminant Analysis
LECTURE 09: DISCRIMINANT ANALYSIS
Spectral Transformation
Review of Statistical Pattern Recognition
ECE – Pattern Recognition Lecture 10 – Nonparametric Density Estimation – k-nearest-neighbor (kNN) Hairong Qi, Gonzalez Family Professor Electrical.
Presentation transcript:

1 Information Extraction Principles for Hyperspectral Data David Landgrebe Professor of Electrical & Computer Engineering Purdue University A Historical Perspective Data and Analysis Factors Hyperspectral Data Characteristics Examples Summary of Key Factors Outline

Sputnik REMOTE SENSING OF THE EARTH Atmosphere - Oceans - Land Brief History National Space Act - NASA formed TIROS I Some 40 Earth Observational Satellites Flown

3 Image Pixels Enlarged 10 Times Thematic Mapper Image

4 Three Generations of Sensors 6-bit data MSS bit data TM bit data Hyperspectral 1986

5 Systems View Sensor On-Board Processing Preprocessing Data Analysis Information Utilization Human Participation with Ancillary Data Ephemeris, Calibration, etc.

6 Scene Effects on Pixel

7 Data Representations Image Space Spectral SpaceFeature Space Sample Image Space - Geographic Orientation Feature Space - For Use in Pattern Analysis Spectral Space - Relate to Physical Basis for Response

8 Data Classes

9 SCATTER PLOT FOR TYPICAL DATA

10 BHATTACHARYYA DISTANCE Mean Difference TermCovariance Term

11 Vegetation in Spectral Space Laboratory Data: Two classes of vegetation

12 Scatter Plots of Reflectance

13 Vegetation in Feature Space

14 Hughes Effect G.F. Hughes, "On the mean accuracy of statistical pattern recognizers," IEEE Trans. Inform. Theory., Vol IT-14, pp , 1968.

15 A Simple Measurement Complexity Example

16 Classifiers of Varying Complexity Quadratic Form Fisher Linear Discriminant - Common class covariance Minimum Distance to Means - Ignores second moment

17 Classifier Complexity - con’t Correlation Classifier Spectral Angle Mapper Matched Filter - Constrained Energy Minimization Other types - “Nonparametric” ê Parzen Window Estimators ê Fuzzy Set - based ê Neural Network implementations ê K Nearest Neighbor - K-NN ê etc.

18 Covariance Coefficients to be Estimated Assume a 5 class problem in 6 dimensions Normal maximum likelihood - estimate coefficients a and b Ignore correlation between bands - estimate coefficients b Ignore correlation between bands - estimate coefficients d Class 1Class 2Class 3Class 4Class 5 bbbbb a ba ba ba ba b a a ba a ba a ba a ba a b a a a b a a a ba a a ba a a b a a a b a a a a ba a a a ba a a a ba a a a ba a a a b a a a a a ba a a a a ba a a a a ba a a a a ba a a a a b Assume common covariance - estimate coefficients c and d Common Covar. d c d c c d c c c d c c c c d c c c c c d

19 EXAMPLE SOURCES OF CLASSIFICATION ERROR

20 Number of Coefficients to be Estimated Assume 5 classes and p features

21 Intuition and Higher Dimensional Space Borsuk’s Conjecture: If you break a stick in two, both pieces are shorter than the original. Keller’s Conjecture: It is possible to use cubes (hypercubes) of equal size to fill an n-dimensional space, leaving no overlaps nor underlaps. Science, Vol. 259, 1 Jan 1993, pp Counter-examples to both have been found for higher dimensional spaces.

22 The Geometry of High Dimensional Space The Volume of a Hypercube concentrates in the corners The Volume of a Hypersphere concentrates in the outer shell

23 Some Implications èHigh dimensional space is mostly empty. Data in high dimensional space is mostly in a lower dimensional structure. èNormally distributed data will have a tendency to concentrate in the tails; Uniformly distributed data will concentrate in the corners.

24 Volume of a hypersphere = How can that be? Differential Volume at r =

25 How can that be? (continued) The Probability Mass at r =

26 MORE ON GEOMETRY The diagonals in high dimensional spaces become nearly orthogonal to all coordinate axes Implication: The projection of any cluster onto any diagonal, e.g., by averaging features could destroy information

27 STILL MORE GEOMETRY The number of labeled samples needed for supervised classification increases rapidly with dimensionality In a specific instance, it has been shown that the samples required for a linear classifier increases linearly, as the square for a quadratic classifier. It has been estimated that the number increases exponentially for a non-parametric classifier. For most high dimensional data sets, lower dimensional linear projections tend to be normal or a combination of normals.

28 A HYPERSPECTRAL DATA ANALYSIS SCHEME 200 Dimensional Data Class Conditional Feature Extraction Feature Selection Classifier/Analyzer Class-Specific Information

29 Finding Optimal Feature Subspaces Feature Selection (FS) Discriminant Analysis Feature Extraction (DAFE) Decision Boundary Feature Extraction (DBFE) Projection Pursuit (PP). Available in MultiSpec via WWW at: Additional documentation via WWW at:

30 Hyperspectral Image of DC Mall HYDICE Airborne System 1208 Scan Lines, 307 Pixels/Scan Line 210 Spectral Bands in µm Region 155 Megabytes of Data (Not yet Geometrically Corrected)

31 Define Desired Classes Training areas designated by polygons outlined in white

32 Thematic Map of DC Mall Legend OperationCPU Time (sec.)Analyst Time Display Image18 Define Classes< 20 min. Feature Extraction12 Reformat67 Initial Classification34 Inspect and Mod. Training ≈ 5 min. Final Classification33 Total164 sec = 2.7 min.≈ 25 min. Roofs Streets Grass Trees Paths Water Shadows (No preprocessing involved)

33 Hyperspectral Potential - Simply Stated Assume 10 bit data in a 100 dimensional space. That is (1024) 100 ≈ discrete locations Even for a data set of 10 6 pixels, the probability of any two pixels lying in the same discrete location is vanishingly small.

34 Summary - Limiting Factors Preprocessing Data Analysis Information Utilization Human Participation with Ancillary Data Sensor On-Board Processing Ephemeris, Calibration, etc. Scene - The most complex and dynamic part Sensor - Also not under analyst’s control Processing System - Analyst’s choices

35 Limiting Factors Scene - Varies from hour to hour and sq. km to sq. km Sensor - Spatial Resolution, Spectral bands, S/N Processing System - Classes to be labeled Number of samples to define the classes Complexity of the Classifier Features to be used - Exhaustive, - Separable, - Informational Value,

36 Source of Ancillary Input Possibilities Ground Observations “Imaging Spectroscopy” - From the Ground - Of the Ground Previously Gather Spectra “End Members” Image Space Spectral Space Feature Space.

37 Use of Ancillary Input A Key Point: Ancillary input is used to label training samples. Training samples are then used to compute class quantitative descriptions Result: This reduces or eliminates the need for many types of preprocessing by normalizing out the difference between class descriptions and the data