[Image Similarity Based on Histogram] Wang wan 581 Project Prof. Longin Jan Latecki.

Slides:



Advertisements
Similar presentations
Point Processing Histograms. Histogram Equalization Histogram equalization is a powerful point processing enhancement technique that seeks to optimize.
Advertisements

Image Segmentation Longin Jan Latecki CIS 601. Image Segmentation Segmentation divides an image into its constituent regions or objects. Segmentation.
With CSS, a color is most often specified by: a HEX value - like "#ff0000" an RGB value - like "rgb(255,0,0)" a color name - like "red“ Example h1.
Announcements Project 2 due today Project 3 out today –demo session at the end of class.
Theoretical Probability Distributions We have talked about the idea of frequency distributions as a way to see what is happening with our data. We have.
Chapter 8 Content-Based Image Retrieval. Query By Keyword: Some textual attributes (keywords) should be maintained for each image. The image can be indexed.
Digital Image Processing Lecture11: Histogram Processing.
6. Gray level enhancement Some of the simplest, yet most useful, image processing operations involve the adjustment of brightness, contrast or colour in.
Bit Depth and Spatial Resolution SIMG-201 Survey of Imaging Science © 2002 CIS/RIT.
Lesson 5 Histograms and Box Plots. Histograms A bar graph that is used to display the frequency of data divided into equal intervals. The bars must be.
T.Sharon 1 Internet Resources Discovery (IRD) Introduction to MMIR.
Original image: 512 pixels by 512 pixels. Probe is the size of 1 pixel. Picture is sampled at every pixel ( samples taken)
Lec 5 Feb 10 Goals: analysis of algorithms (continued) O notation summation formulas maximum subsequence sum problem (Chapter 2) three algorithms image.
B a c kn e x t h o m e Classification of Variables Discrete Numerical Variable A variable that produces a response that comes from a counting process.
Copyright © Cengage Learning. All rights reserved. 5 Integrals.
Business Statistics: A First Course, 5e © 2009 Prentice-Hall, Inc. Chap 6-1 Chapter 6 The Normal Distribution Business Statistics: A First Course 5 th.
How Images are Represented Bitmap images (Dots used to draw the image) Monochrome images 8 bit grey scale images 24 bit colour Colour lookup tables Vector.
1 Activity and Motion Detection in Videos Longin Jan Latecki and Roland Miezianko, Temple University Dragoljub Pokrajac, Delaware State University Dover,
Area Area is the amount of surface space that a flat object has.
UNIT FOUR/CHAPTER NINE “SAMPLING DISTRIBUTIONS”. (1) “Sampling Distribution of Sample Means” > When we take repeated samples and calculate from each one,
Image Segmentation CIS 601 Fall 2004 Longin Jan Latecki.
Image segmentation by clustering in the color space CIS581 Final Project Student: Qifang Xu Advisor: Dr. Longin Jan Latecki.
13.1 day 2 level curves. Level curves One way to visualize a function of two variables is to use a scalar field in which the scalar z = f(x,y) is assigned.
Similar Figures 4-3 Problem of the Day A rectangle that is 10 in. wide and 8 in. long is the same shape as one that is 8 in. wide and x in. long. What.
Today we will derive and use the formula for the area of a triangle by comparing it with the formula for the area of a rectangle. derive = obtain or receive.
Finding Area of Other Polygons
Chapter 2 Arithmetic Strategies & Area
Write the standard number 38,440 in scientific notation. A × 10 6 B × 10 3 C × 10 4 D ×
BACKGROUND LEARNING AND LETTER DETECTION USING TEXTURE WITH PRINCIPAL COMPONENT ANALYSIS (PCA) CIS 601 PROJECT SUMIT BASU FALL 2004.
Copyright © Cengage Learning. All rights reserved. 2 Descriptive Analysis and Presentation of Single-Variable Data.
Are You Smarter Than a 5 th Grader?. 1,000,000 5th Grade Topic 15th Grade Topic 24th Grade Topic 34th Grade Topic 43rd Grade Topic 53rd Grade Topic 62nd.
Chapter 2 Describing Data.
Color and Resolution Introduction to Digital Imaging.
Evaluating Algebraic Expressions 5-5 Similar Figures Preparation for MG1.2 Construct and read drawings and models made to scale. California Standards.
Copyright © Cengage Learning. All rights reserved. 4 Integrals.
Image Similarity Longin Jan Latecki CIS Dept. Temple Univ., Philadelphia
Image segmentation Prof. Noah Snavely CS1114
September 17, 2013Computer Vision Lecture 5: Image Filtering 1ColorRGB HSI.
CIS 601 – 04 Image ENHANCEMENT in the SPATIAL DOMAIN Longin Jan Latecki Based on Slides by Dr. Rolf Lakaemper.
Math 3033 Wanwisa Smith 1 Base on text book: A Modern Introduction to Probability and Statistics Understanding Why and How By: F.M. Dekking, C. Kraaikamp,
Clustering Prof. Ramin Zabih
1 Motion Analysis using Optical flow CIS601 Longin Jan Latecki Fall 2003 CIS Dept of Temple University.
1 Copyright © Cengage Learning. All rights reserved. 3 Functions and Graphs 3.1Rectangular Coordinate Systems.
Image Segmentation by Histogram Thresholding Venugopal Rajagopal CIS 581 Instructor: Longin Jan Latecki.
CY1B2 Statistics1 (ii) Poisson distribution The Poisson distribution resembles the binomial distribution if the probability of an accident is very small.
COMPUTER GRAPHICS. Once you scan an image or take a picture with your digital camera, it becomes digitized. Made up of hundreds of thousands of pixels.
Content-Based Image Retrieval QBIC Homepage The State Hermitage Museum db2www/qbicSearch.mac/qbic?selLang=English.
CS 325 Introduction to Computer Graphics 04 / 12 / 2010 Instructor: Michael Eckmann.
Content-Based Image Retrieval (CBIR) By: Victor Makarenkov Michael Marcovich Noam Shemesh.
Course14 Dynamic Vision. Biological vision can cope with changing world Moving and changing objects Change illumination Change View-point.
A parallelogram with opposite equal acute and obtuse angles and four equal sides. Diagonals 4 equal sides.
Review of Statistical Terms Population Sample Parameter Statistic.
Chapter 1: Image processing and computer vision Introduction
PART TWO Electronic Color & RGB values 1. Electronic Color Computer Monitors: Use light in 3 colors to create images on the screen Monitors use RED, GREEN,
Toddrick’s Shape Scrapbook By:Toddrick Newton. The perimeter, P, of a rectangle is given by the formula P = 2(l + w) where l is the length width of the.
Similar Triangles Triangles that have the same shape but not necessarily the same size. Corresponding angles are congruent. Meaning they have the same.
Image Similarity Presented By: Ronak Patel Guided By: Dr. Longin Jan Latecki.
Statistical Fundamentals: Using Microsoft Excel for Univariate and Bivariate Analysis Alfred P. Rovai Histograms PowerPoint Prepared by Alfred P. Rovai.
Graphics Basic Concepts 1.  A graphic is an image or visual representation of an object.  A visual representation such as a photo, illustration or diagram.
STRETCHES AND SHEARS.
Copyright © Cengage Learning. All rights reserved. 4 Integrals.
The Normal Distribution: Comparing Apples and Oranges
Ch. 4 – Displaying Quantitative Data (Day 1)
Fall 2012 Longin Jan Latecki
Vectors.
#31 He was seized by the fidgets,
2.3 Estimating PDFs and PDF Parameters
Digital Image Processing
CIS Dept. Temple Univ., Philadelphia
CIS Dept. Temple Univ., Philadelphia
Presentation transcript:

[Image Similarity Based on Histogram] Wang wan 581 Project Prof. Longin Jan Latecki

Image Similarity  Pixel based  Histogram based  Shape similarity  Etc.

What is histogram? (3, 8, 5)

Image Histogram  If f:[1, n]x[1, m]  [0, 255] is a gray value image, then H(f): [0, 255]  [0, n*m] is its histogram, where H(f)(k) is the number of pixels (i, j) such that f(i, j)=k  If f is a RGB image. then we compute Hr(f),Hg(f),Hb(f) seperately  Example Example

A rgb image I

Hr Hg

Hb

Real Process  Read and resize all the images into 4-D array ims.  Use imhist to compute histograms  Compute the difference of 2 images using the formulasCompute the difference of 2 images using the formulas  Display the compare_order graphDisplay the compare_order graph

for k=1:filenum imk=ims(:,:,:,k); res(1,k)=imdiff1(imi,imk,bins); res(2,k)=imdiff2(imi,imk,bins); res(3,k)=imdiff3(imi,imk,bins); res(4,k)=imdiff4(imi,imk,bins); res(5,k)=imdiff4(imi,imk,bins); end; for method_k = 1:5 [y,ind(method_k,:)]=sort(res(method_k,:)); for file_k = 1:filenum ord(method_k,ind(method_k,file_k))=file_k; end; figure; end; figure; for k = 1:5 subplot(2,3,k); plot(ord(k,:)); title(strcat('method ',int2str(k))); end; Display the compare_order graph  using function sort and plot

Formulas: a,b are two images (1) (2) (3) For gray-value image, (1) and (3) get the same result. For rgb image, they are a little different.

(4) statistic formula For pixel_based, this formula get best result. For hist_based, this formula get worst result.(for hist_based method,

(5) Let v(a) be a vector of all c(a,x,y) values assigned to all pixels in the image a. Image similarity can be expressed as normalized inner products of such vectors. Since it yields maximum values for equal frames, a possible disparity measure is

Image Database Used  Database1: Consists of 100 images. 100 shots taken from four different movies Images 1-40 : Mr. Bean's Christmas Images 41-70: S hots from House Tour Images 71:90: Shots of squares on a desk Images 91:100: S hots from a Kylie Minogue interview

 Database3: Consists of 10 images. Images 0-4: image of ocean scenes. in fact, they are same image with different rotating angles Images 5-6: other images of ocean scenes  Database2: Consists of 30 images. Images 0-9 are images of rose Images are images with red objects other than rose Images are ocean scenes Image are images of tiger

Result & Analysis  1. Hist_based image similarity costs much less time than pixel_based imge similarity. Suppose the width of image is w, height is h. Suppose we compute the histgram of the image into m bins For every formula, if the time costs are O(n), then hist_based costs O(m), pixel_based costs O(w*h). In practice, when we compute the pic_set1 with formula (1): w=80, h=112, m=30; time of pixel_based method: O( ) time of hist_based method: O(900)

 2. Bins of histogram can be less than 255. If we divide the histogram into 50 pins, the result is almost the same as we divide the histogram into 255 bins. We can even divide the histogram into 20 pins, the result is much the same. It doesn’t mean that the higher the number of bin is,the better result is.  3. For Pixel_based image similarity test In gerneal, the formula (4) get the best result. The most typical case is in pic_set3, it get result as good as hist_based methods. The formula (5) get the worst result. (4)>(3),(1)>(2)>(5)  4. For hist_based image similarity test In gerneal, the formula (4) get the worst result. (3),(1)>(5) >(2) >(4)

 5. In pic_set1 and pic_set3, hist_based methods get better result than pixel_based methods. Conclusion: when different pictures have the same objects but different motion/ position, the hist_based method is better than pixel_based methods.  6. In pic_set2, pixel _based methods get better result than hist_based methods. Conclusion: when different pictures have different objects with the same color, in general, hist_based is worse than the pixel_based methods