Skip to main content

Banery wysuwane

liternicza seryjna okładka w pomarańczowym kolorze
Analysis of User Criteria and Backward MCDM Methods with Applications to Content-Based Image Retrieval and Assessment

Author
Product category
Nauki techniczne » Informatyka
ISBN
978-83-7464-697-0
ISSN
0867-6631
Publication type
monografia
Format
B5
Binding
miękka
Number of pages
161
Publication date
2012
Description

The main thrust of this monograph is related to the problem of the expression of queries by users, who are not usually able to provide visual examples or to express their graphic query in a form appropriate for automatic interpretation. On the other hand, users may be able to indicate examples of images that are similar (or dissimilar) to what they are looking for. This form of providing information about the query is called Relevance Feedback (RF). A visual search engine should, therefore, include an algorithm capable of generalizing a user’s preferences based on his assessment of the relevance of sample images. For applications where image retrieval can be based on low-level features of images, such as colour, texture and shape, we have proposed an algorithm with an approximation of user preferences using an RBF neural network. This approach is adequate for searching in a specified class of images that contain single objects distinct from the background. An example application is the interactive atlas of species that we have developed to test our method.


Główny nurt niniejszej monografii dotyczy problemu formułowania zapytania przez użytkowników, którzy zazwyczaj nie są w stanie dostarczyć graficznych przykładów poszukiwanych obrazów ani wyrazić zapytania graficznego w formie zdatnej do automatycznej interpretacji. Tak więc wyszukiwarka graficzna powinna zawierać algorytm uogólniający preferencje użytkownika na podstawie określenia przez niego, w jakim stopniu przykładowe obrazy spełniają jego oczekiwania (ang. relevance feedback). Dla zastosowań w których wyszukiwanie może być oparte na cechach niskiego poziomu, jak kolor, tekstura czy kształt, zaproponowaliśmy algorytm z aproksymacją preferencji użytkownika przy użyciu sieci neuronowej typu RBF. Metoda ta przewidziana jest do wyszukiwania obrazów należących do określonej klasy, zawierających pojedyncze obiekty wyodrębnione z tła. Przykładowym zastosowaniem jest interaktywny atlas gatunków, który powstał w celu przetestowania zaproponowanej metody.


Wydawnictwa nie prowadzą sprzedaży książek z serii "Rozprawy Monografie". Zainteresowanych prosimy o kontakt z ich autorami. 

Contents

Summary  9
Streszczenie  11
Acknowledgments  13
Notation  15
Preface  17
1. State of the art in Content-Based Image Retrieval  20
1.1. Surface descriptors  20
1.1.1. Colour descriptors  21
1.1.2. Ensuring colour constancy in colour-based retrieval  22
1.1.3. Texture descriptors  23
1.1.4. Similarity measures for surface descriptors  25
1.2. Keypoint descriptors and bag-of-features approach  25
1.3. Shape-based retrieval  27
1.3.1. Methods for shape-based retrieval – 2D approach  27
1.3.1.1. Basic region-based features  27
1.3.1.2. Curvature scale space  29
1.3.1.3. Retrieval by alignment  29
1.3.2. Methods for shape-based retrieval – 3D approach  30
1.3.2.1. Feature-based methods  31
1.3.2.2. Graph-based methods 32
1.3.2.3. Geometry-based methods  33
1.3.2.4. Creating 3D models from real objects  33
1.4. Relevance feedback: virtual query and distance-based retrieval  35
1.4.1. Interaction with the user and virtual queries  35
1.4.2. Rocchio formula  36
1.4.3. Combining features in the virtual query scheme  37
1.4.4. Shortcomings of CBIR based on virtual queries  38 
2. Shape retrieval by alignment  39
2.1. Introduction to the Hausdorff distance  39
2.2. Formulation of recognition and retrieval as a multilevel optimisation problem  41
2.3. Efficient calculation of the Hausdorff distance  42
2.3.1. State of the art in increasing efficiency of the HD calculation  43
2.3.2. Contour approximation at the level L1  45
2.3.3. Contour approximation at the level L2  46
2.3.4. Contour pruning  49
2.3.4.1. Mathematical basis of the proposed contour pruning method  49
2.3.4.2. An application to contour recognition  50
2.3.4.3. Efficiency  51
2.3.4.4. Experimental time comparison  52
2.3.5. Optimisation in the transformation space  54
2.3.6. Database navigation and pruning  55
2.3.6.1. Using the triangle inequality for determining the search order  55
2.3.6.2. Using additional image features to increase the efficiency of the HD evaluation algorithms  57
2.3.6.3. Experimental time comparison  57
2.3.7. Summary of inaccuracies for the presented speeding-up methods  58
2.4. A practical application: 3D object retrieval by shape alignment  59
2.4.1. 3D modelling vs. direct silhouette matching  59
2.4.2. Dissimilarity measures  61
2.4.3. Implementation  63
3. Retrieval of non-homogeneous objects with preference approximation in feature spaces  66
3.1. General outline of matching scheme and selection of descriptors  66
3.2. Adaptation of distance-based method for complex objects  68
3.3. Approximation of user preferences by RBF  70
3.4. Experimental comparison of algorithms for non-homogeneous object retrieval  72
3.5. Conclusions  74
4. Elicitation of relevant features based on relational MCDM  75
4.1. Relevance feedback by pairwise comparisons  79
4.1.1. Pairwise comparisons and the Analytic Hierarchy Process  79
4.1.2. Basic AHP algorithm  80
4.1.3. The proposed algorithm for information retrieval  82
4.1.4. A practical application: an image retrieval system  87
4.1.5. Performance of the method  88
4.1.6. Conclusions  91
4.2. Relevance feedback by graph of relations  92
4.2.1. The concept of user criteria retrieval based on graph of relations  92
4.2.2. Criteria elicitation and information retrieval based on ELECTRE methodology  93
4.2.2.1. The ELECTRE III method  94
4.2.2.2. Proposed method for criteria elicitation  96
4.2.3. Application to Content-Based Image Retrieval  99 
4.2.4. Performance and efficiency  103
4.3. Relevance feedback by individual assessment  105
4.3.1. Reference sets  105
4.3.2. The proposed algorithm for the criteria selection  106
4.3.3. An example of real-life application and performance assessment  108
4.4. Comparison of the performance and efficiency  113
5. Application of multicriteria image analysis and relevance feedback for the glass melting process control  116
5.1. Preliminary image processing: segmentation and mapping  117
5.1.1. Batch segmentation  119
5.1.2. Lens sediment segmentation  120
5.1.3. Image mapping  120
5.2. Analysis of temperature symmetry  122
5.2.1. Batch blanket asymmetry indicator  122
5.2.2. Calculation of CCT  124
5.2.3. The influence of reversals on temperature distribution asymmetry  126
5.2.4. Areas of batch symmetry and glass symmetry  127
5.2.5. Cross indicators of temperature asymmetry  128
5.2.6. Inferring about process settings from the asymmetry indicators  130
5.2.7. An example of the analysis of melting symmetry  131
5.3. Elicitation of melting criteria based on pairwise comparisons  135
5.3.1. Preliminary image processing and the calculation of potential criteria  136
5.3.2. The backward AHP and its application in the elicitation of relevant criteria  137
5.3.3. Elicitation of relevant parameters  141
5.4. Summary of the results and conclusions  144
6. Final conclusions  147
References  149

Contents
Price
0.00
In order to arrange international shipping details and cost please contact wydawnictwa@agh.edu.pl