Title: Graphbased approaches in image analysis: a review
 1Graph-based approaches in image analysis a review
ANR Project 
Navidomass 
- Salim Jouili 
 - Supervisor 
 - S.A. Tabbone 
 
QGAR  LORIA Nancy
Réunion Navidomass Paris, le 21 Mars 2008 
 2Outline 
- Introduction 
 - Graph-based representation 
 - Similarity measures of graphs 
 - Edit distance 
 - Papadopolous and Manolopoulos measure 
 - Maximal common Subgraph 
 - Graph probing 
 - Median Graph 
 - Applications 
 - Conclusion 
 
  3introduction
- Powerful structured-based representation 
 - Used with flexibility in processing of a large 
variety of images types (the ancient documents, 
the electric and architectural plans, natural 
images, medical images...).  - Preserves topographic information of the image as 
well as the relationship between the components.  - In the two last decades many works have been 
developed.  - Step in very subfield of image analysis  
 - Pattern Recognition 
 - Segmentation 
 - CBIR (Content-based image retrieval)
 
  4Graph-based representation
- Bunke ,PAMI82 1 
 - (x,y)  vertices attributes 
 - 1,2 and 3  vertices labels 
 - 1 Final point 
 - 2 angle 
 - 3  T intersection
 
  5Graph-based representation
Multilayer segmentation Homogeneous zones 
 6Graph-based representation
- Region adjacency Graphs 
 - Fauqueur, PhD 2003 3 
 
Original image
a RAG Representation Of the segmented image 
 7Graph-based representation
- Region adjacency Graphs 
 - Llados, PAMI01 4 
 - Extraction regions of a plane graph by Jiang and 
Bunke algorithm 5.  
V1
e1
V2
R1
e8
e2
V3
V6
R2
e7
e3
e4
e6
V5
V4
e5
- A RAG G 
 - Vertices represent the regions in G 
 - Edges  represent the regions adjacency in G
 
R3
A plane Graph G representing line drawing 
 8Graph-based representation
- GCap Graph-based Automatic Image Captioning, J. 
Pan, MDDE04 6.  
  9Aims of graph-based representation
- Most of works in graph-based representation, 
notably in document analysis, sought some 
resemblance measures between represented objects 
in order to   - Classify 
 - Match 
 - Index 
 - ... 
 
  10Similarity measures for graphs
- Edit distance 
 - Maximal common subgraph (MCS) 
 
1 operation Edge deletion
1 operation Vertex Substitution
G1
G2
D(G1,G2)  2
G1
G2
Dmcs(G1,G2)  1- (3/4)0.25 
 11Similarity measures for graphs
- Papadoupolos and Manolopoulos Measure 7 
 
- Sorted graph histogram  
 - SH 1 V5(3), V4(3), V1(3), V6(2), V3(2), V2(1)
 
V2
V1
V3
- Sorted graph histogram  
 - SH 2  V4(4), V3(4), V1(4), V6(3), V5(3), V2(2)
 
V5
V6
V4
V2
V1
Dpa.  Mano(G1,G2) L1(SH1,SH2)6
V3
V4
Primitive operations are  vertex insertion , 
vertex deletion and vertex update
V6
V5 
 12Similarity measures for graphs
- Graph Probing, Lopresti, IJDAR2004 8 
 - How many vertices with degree n are present in 
graph G (V,E)? PR collect the response from the 
graphs  - PR(G)  (n0,n1,n2,) where niv?V deg(v) i
 
Dprobing(G1,G2) L1(PR(G1),PG(G2) 
 13Median Graph
- The generalized median graph aims to extract 
essential information from a whole of set of 
graphs in only one prototype 
The generalized median graph
A set of graphs 
 14Median Graph
- GGM  arg ming?U?i1 d(g,gi) 
 - Where U is the set of all the graphs that can be 
built from the original set of graphs.  - Jiang Propose a genetic algorithm, GbR99 9 
 - Hlaoui proposed a solution based on the 
decomposition of the problem of minimizing the 
sum of distances in two parts, nodes and edges. 
GbR03 10 
  15Applications
- Content-based image retrieval  
 - Berretti proposed a technique of graph matching 
and indexing dedicated to the graph-models in 
content-based retrieve. Using m-tree indexing 
method. PAMI2001 11.  - Segmention 
 - Felzenszwalb proposed a complete graph-based 
approach for the segmentation of colour images. 
12  - ... 
 
  16conclusion
- Graph-based representation  flexible, universal 
(documents type), spatial information.  - Useful in many field in image analysis. 
 - Many solution in measurement of similarity 
between graphs ? depends from the data stored in 
graphs.  - Ambitious research field notably for 
Content-based image retrieval. 
  17REFERENCES
- 1 H. Bunke. Attributed of programmed graph 
grammars and their application to schematic 
diagram interpretation. IEEE Transactions on 
Pattern Analysis and Machine Intelligence, 4(6), 
Novembre 1982.  - 2 A. Karray. Recherche de lettrines par le 
contenu. Master's thesis, Laboratoire L3i, 
Universités de La Rochelle et de Sfax, France et 
Tunisie, 2006.  - 3 J. Fauqueur. Contributions pour la Recherche 
d'Images par Composantes Visuelles. PhD thesis, 
INRIA -Université Versailles St Quentin, 2003.  - 4 J. Lladòs, E. Martí, and J. J. Villanueva. 
Symbol recognition by error-tolerant subgraph 
matching betweenregion adjacency graphs. IEEE 
Transactions on Pattern Analysis and Machine 
Intelligence, 23(10),2001.  - 5 Jiang, X.Y., Bunke, H., An Optimal Algorithm 
for Extracting the Regions of a Plane Graph, 
Pattern Recognition Letters (14), 1993, pp. 
553-558.  - 6 J. Pan, H.Yang, C. Faloutsos, and P. Duygulu. 
Gcap  Graph-based automatic image captioning. In 
Proceedings of the 4th International Workshop on 
Multimedia Data and Document Engineering, 2004.  - 7 A. N. Papadopoulos and Y. Manolopoulos. 
Structure-based similarity search with graph 
histograms. Proceedings of International Workshop 
on Similarity Search (DEXA IWOSS'99), pages 
174178, Septembre 1999.  - 8 D. Lopresti and G. Wilfong. A fast technique 
for comparing graph representations with 
applications to perform evaluation. IJDAR, 
6219229, 2004.  - 9 X. Jiang, A. Munger, and H. Bunke. Scomputing 
the generalized median of a set of graphs. 2nd 
IAPR-TC-IS Workshop on Graph Based 
Representations.  - 10 A. Hlaoui and S.Wang. A new median graph 
algorithm. IAPR Workshop on GbRPR, LNCS 2726, 
pages 225234, 2003.  - 11 S. Berretti, A. D. Bimbo, and E. Vicario. 
Efficient matching and indexing of graph models 
in content-based retrieval. IEEE Transactions on 
Pattern Analysis and Machine Intelligence, 
23(10)10891105, 2001.  - 12 P. F. Felzenszwalb and D. P. Huttenlocher. 
Efficient graph-based image segmentation. 
International Journal of Computer Vision, 59(2), 
Septembre 2004.