Image Retrieval Based on Fractal Signatures John Y. Chiang Z. Z. Tsai

Slides:



Advertisements
Similar presentations
8.2 Kernel And Range.
Advertisements

3D Geometry for Computer Graphics
Wavelets Fast Multiresolution Image Querying Jacobs et.al. SIGGRAPH95.
Recovery of affine and metric properties from images in 2D Projective space Ko Dae-Won.
3D Shape Histograms for Similarity Search and Classification in Spatial Databases. Mihael Ankerst,Gabi Kastenmuller, Hans-Peter-Kriegel,Thomas Seidl Univ.
Mapping: Scaling Rotation Translation Warp
Chapter 5 Orthogonality
Computer Graphics Recitation 2. 2 The plan today Learn about rotations in 2D and 3D. Representing rotations by quaternions.
3-D Geometry.
A new predictive search area approach for fast block motion estimation Kuo-Liang Chung ( 鍾國亮 ) Lung-Chun Chang ( 張隆君 ) 國立台灣科技大學資訊工程系暨研究所 IEEE TRANSACTIONS.
中華大學 資訊工程系 Fall 2002 Chap 4 Laplace Transform. Page 2 Outline Basic Concepts Laplace Transform Definition, Theorems, Formula Inverse Laplace Transform.
6 1 Linear Transformations. 6 2 Hopfield Network Questions.
Orthogonality and Least Squares
Chap 1 First-Order Differential Equations
Transforms: Basis to Basis Normal Basis Hadamard Basis Basis functions Method to find coefficients (“Transform”) Inverse Transform.
University of Texas at Austin CS395T - Advanced Image Synthesis Spring 2006 Don Fussell Orthogonal Functions and Fourier Series.
1 Faculty of Information Technology Generic Fourier Descriptor for Shape-based Image Retrieval Dengsheng Zhang, Guojun Lu Gippsland School of Comp. & Info.
BACKGROUND LEARNING AND LETTER DETECTION USING TEXTURE WITH PRINCIPAL COMPONENT ANALYSIS (PCA) CIS 601 PROJECT SUMIT BASU FALL 2004.
Digital Image Processing, 3rd ed. © 1992–2008 R. C. Gonzalez & R. E. Woods Gonzalez & Woods Matrices and Vectors Objective.
南台科技大學 資訊工程系 Automatic Website Summarization by Image Content: A Case Study with Logo and Trademark Images Evdoxios Baratis, Euripides G.M. Petrakis, Member,
Unit 5: Geometric Transformations.
Geometric Transformations:
AN ORTHOGONAL PROJECTION
H.264/AVC 基於影像複雜度與提早結束之快速 階層運動估計方法 Content-Based Hierarchical Fast Motion Estimation with Early Termination in H.264/AVC 研究生:何銘哲 指導教授:蔣依吾博士 中山大學資訊工程學系.
Chapter 9 Transformations.
Transformations. Congruent Similar Image vs Pre-image Pre-image: the figure before a transformation is applied. Image: The figure resulting from a transformation.
9.1 Translations -Transformation: a change in the position, shape, or size of a geometric figure -Preimage: the original figure -Image: the resulting figure.
Pre-AP Unit 7 Vocabulary. Area – the measurement attribute that describes the number of square units a figure or region covers Circumference – a linear.
6.7.1 Perform Similarity Transformations.  Remember previous we talked about 3 types of CONGRUENCE transformations, in other words, the transformations.
2005/12/021 Content-Based Image Retrieval Using Grey Relational Analysis Dept. of Computer Engineering Tatung University Presenter: Tienwei Tsai ( 蔡殿偉.
VECTORS (Ch. 12) Vectors in the plane Definition: A vector v in the Cartesian plane is an ordered pair of real numbers:  a,b . We write v =  a,b  and.
NTIT1 A chaos-based robust wavelet- domain watermarking algorithm Source: Chaos, Solitions and Fractals, Vol. 22, 2004, pp Authors: Zhao Dawei,
2005/12/021 Fast Image Retrieval Using Low Frequency DCT Coefficients Dept. of Computer Engineering Tatung University Presenter: Yo-Ping Huang ( 黃有評 )
Lecture 13 Inner Product Space & Linear Transformation Last Time - Orthonormal Bases:Gram-Schmidt Process - Mathematical Models and Least Square Analysis.
1 Chapter 2: Geometric Camera Models Objective: Formulate the geometrical relationships between image and scene measurements Scene: a 3-D function, g(x,y,z)
Content-Based Image Retrieval QBIC Homepage The State Hermitage Museum db2www/qbicSearch.mac/qbic?selLang=English.
Intelligent Database Systems Lab 國立雲林科技大學 National Yunlin University of Science and Technology 1 Region-based image retrieval using integrated color, shape,
Lecture 17 - Approximation Methods CVEN 302 July 17, 2002.
Signal & Weight Vector Spaces
6 6.3 © 2016 Pearson Education, Inc. Orthogonality and Least Squares ORTHOGONAL PROJECTIONS.
CHAPTER Four Linear Transformations. Outlines Definition and Examples Matrix Representation of linear transformation Similarity.
2008/7/10 1 以空間關係相鄰圖為基礎之空間關 係相似性量測方法 Retrieval by spatial similarity based on interval neighbor group 研究生:黃彥人 指導教授:蔣依吾博士 中山大學資訊工程學系.
Chapter 9 Properties of Transformations Warren Luo Matthew Yom.
CHAPTER 4 TRANSFORMATIONS  What you will learn:  Perform translations  Perform compositions  Solve real-life problems involving compositions 4.1.
1 Objective To provide background material in support of topics in Digital Image Processing that are based on matrices and/or vectors. Review Matrices.
Chapter 5 Eigenvalues and Eigenvectors
Singular Value Decomposition and its applications
8.2 Kernel And Range.
Discrete Fourier Transform (DFT)
Do Now.
隱 像 術 張真誠 國立中正大學資訊工程研究所.
Review: Transformations
Transformations Chapter 4.
Math II: Unit 1: Transformations
Wavelets : Introduction and Examples
Matrices and Vectors Review Objective
Y. Davis Geometry Notes Chapter 9.
Self-Organizing Maps for Content-Based Image Database Retrieval
Transformations Sections
9.1 Translations -Transformation: a change in the position, shape, or size of a geometric figure -Preimage: the original figure -Image: the resulting figure.
Objective To provide background material in support of topics in Digital Image Processing that are based on matrices and/or vectors.
Theorems about LINEAR MAPPINGS.
Day 84 – Enlargement.
Fast color texture recognition using chromaticity moments
Lecture 13: Singular Value Decomposition (SVD)
Elementary Linear Algebra Anton & Rorres, 9th Edition
1.) Create a design (image) on the graph paper with 3 vertices land on whole number (integer) coordinates, in the upper left quadrant (Quadrant II). Label.
Transformations.
Day 47 – Fixed points and invertibility in rigid motion
Presentation transcript:

Image Retrieval Based on Fractal Signatures John Y. Chiang Z. Z. Tsai 蔣依吾 國立中山大學 資訊工程學系 E-mail: chiang@cse.nsysu.edu.tw Web: http://image.cse.nsysu.edu.tw

Block Diagram of Image retrieval System

Feature used for querying Text, color, shape, texture , spatial relations Image Retrieval Systems: IBM QBIC: color , texture, shape Berkeley BlobWorld: color, texture, location and shape of region (blobs) Columbia VisualSEEK: text, color , texture, shape, spatial relations

Ideal image retrieval 相似之影像有相似之索引檔 不相似之影像有不相似之索引檔 (a) 高相關度影像資訊有高相關索引檔 (b) 索引檔相關度高,其影像資訊相關度高 不相似之影像有不相似之索引檔 (c) 索引檔相關度低,其影像資訊相關度低 (d) 影像資訊相關度低,其索引檔相關度低 Based on Fractal Orthogonal Basis, image index satisfy the four properties.

Fractal Orthogonal Basis Vines, Nonlinear address maps in a one-dimensional fractal model, IEEE Trans. Signal Processing,1993. Training Orthonormal basis vectors. Gram-Schmidt method Compression: each image range block (R) decomposed into a linear combination, Using linear combination coefficient as a feature vector。

Proof of Theorem theorem 1: Let f , g is distance space(K,d)functions, Range block length k,f= , g= , i,j 1.. xi‧xj=0 and | xi |=1 , S={s1s2.. } f has fixed point a,S: query image a’s index, T={t1t2… }, g has fixed point b , T: database image b’s index, if a , b is close, then S , T is close; if S , T is close, then a , b is close。

(a) Proof : a , b is close , S , T is close Proof of Theorem 1 (a) Proof : a , b is close , S , T is close a , b is close , then ||a-b|| thus a , b is close , S , T is close

(b) Proof: S , T is close , a , b is close Proof of Theorem 1 (b) Proof: S , T is close , a , b is close S , T is close , then thus S , T is close , a , b is close

Proof of Theorem theorem 2: Let f , g is distance space(K,d)functions, Range block length k,f= , g= , i,j 1.. xi‧xj=0 and | xi |=1 , S={s1s2.. } f has fixed point a,S: query image a’s index, T={t1t2… }, g has fixed point b , T: database image b’s index, if a , b is not close, then S , T is not close; if S , T is not close, then a , b is not close。

(c) Proof : a , b is not close , S , T is not close Proof of Theorem 2 (c) Proof : a , b is not close , S , T is not close by theorem 1, a , b is close S , T is close thus a , b is not close , S , T is not close (d) Proof: S , T is not close , a , b is not close by theorem 1, a , b is close S , T is close thus S , T is not close , a , b is not close

Advantages based on Fractal Orthogonal Basis Compressed data can be used directly as indices for query. Image index satisfy the four properties. Similarity measurement is easy.

Orthonormal Basis Vectors (a) (b) (c) Figure 3. The 64 fractal orthonormal basis vectors of (a) R, (b) G, and (c) B color components, respectively, derived from an ensemble of 100 butterfly database images. The size of each vector is enlarged by two for ease of observation..

Fourier transform vs Fractal Orthonormal basis

Original image Compressed image

Experimental Results Image database The source of images: http://www.thais.it/entomologia/ http://turing.csie.ntu.edu.tw/ncnudlm/index.html http://www.ogphoto.com/index.html http://yuri.owes.tnc.edu.tw/gallery/butterfly http://www.mesc.usgs.gov/resources/education/butterfly/ http://mamba.bio.uci.edu/~pjbryant/biodiv/bflyplnt.htm Image size:320x240 Total of images:1013 Range block size:8x8 Domain block size:8x8 Domain block set: 64 domain blocks in each color plane

Figure 4. An image retrieval example Figure 4. An image retrieval example. The features from R, G, B color components and brightness level of the query image are all selected. The rectangular area in the upper right-hand corner provides an enlarged viewing window for the image retrieved.

Figure 6. Retrieval results based on a sub-region of a query image with scaling factors 0.8 through 1.2 and rotation angles every 30 degrees.

Future work 1. With Multiple Instance Learning, finding “ideal” feature vector, as query feature. 2. Considering spatial relations of sub-regions. 3. Increasing performance.