Download as ppt, pdf, or txt
Download as ppt, pdf, or txt
You are on page 1of 12

OBJECT RECOGNITION THROUGH

TOPO-GEOMETRIC SHAPE MODELS


USING ERROR-TOLERANT SUBGRAPH
ISOMORPHISMS

Project Guide Presented by


Dr. O. Umamaheshwari H. Krishna Kumar

(2009236010)
ABSTRACT
We propose a method for Shape recognition
based on subgraph isomorphisms, by extracting
topological and geometric properties of a shape in
the shape model , called as topo-geometric shape
model(TGSM). Test graph is recognized as a shape
that yields the largest sub graph isomorphism with
minimal cost of edit operations. Skeletonization is
major part, in which an object in the image is
represented in the form of skeleton, thereby
reducing the memory usage to retrieve the image
and also used to compare the images in a short time.
LITERATURE REVIEW
Reeb graph model:
 It does not achieves invariance to rigid body
transformations by using distance function.
 Difficult to capture rich geometric attributes in
the form of a geometric model.
TGSM:
 This model avoids the above said difficulties.
 The theme of this model is the skeletonization
technique.
Skeletonization Technique
The skeleton S(A) of the set A in the
image can be calculated by splitting the set A
into K iterations, which is known as Sk (A). This is
represented as
K
S(A) = U Sk(A)
k=0
 
With Sk(A) = (A - kB) – ( A - kB ). B
Where B is the structuring element
Need for skeletonization
• To reduce the amount of data required to be
processed.
• To reduce the time required to be processed.
• Extraction of critical features such as end-points,
junction-points, and
connection among the components.
Applications of skeletonization
• Handwritten and printed characters.
• Fingerprint patterns
• Chromosomes & biological cell structures
BLOCK DIAGRAM

NORMALIZING APPLYING
TEST SHAPE
& PLACING SKELETONIZATION
IMAGE EXTRACTION
ALGORITHM

BEST
COMPARISON
MATCHED
OUTPUTS

SUBGRAPH
MODELS
PROCEDURE
 Taking the test image from which we need to
identify the nature of sub image(graph).
 Performing sub image isolation mechanism from
the input test image.
 Normalizing and placing the sub image in the
centre of the test image, so that it will be easy to
apply skeletonization algorithm.
 Comparing the test sub image(graph) with the
desired sub image(graph) models to obtain error
tolerant outputs.
 Several matches should be prioritized based
on comparison.
Results
Test image and its skeleton
Model image and its skeleton
Comparison
Conclusion and future work
By skeletonizing object in the image,We may
easily compare the desired object which is
skeletonized and assumed as test graph with the
input by drawing skeleton for it. Thus comparing
based on skeletons may reduce the comparison time
and it also useful for fast data transfer. This
technique also provides error tolerant output, i.e.
while comparing, even if exact matches are not
present, this technique will provide output by finding
correlation among the objects in the respective
images.
REFERENCES
• Shape matching and object recognition using
shape contexts
• Image inpainting technique to smoothen the
missing objects in the image.

You might also like