Basic of Pattern recognition

1. WELCOME EVERYONE
Good Afternoon

2. Pattern Recognition
Swarnava Sen
MCKV Institute of Engineering, Liluah
IT/2014/028

3. What is pattern ?
o A pattern is an abstract object, or a set of measurements
describing a physical object.

4. What is a pattern CLass?
o A pattern class (or category) is a set of patterns sharing common
attributes.
o A collection of “Similar” (not necessarily identical) objects.
o During recognition given objects are assigned to prescribed
classes.

5. What is pattern
reCOGnitiOn ?
o Theory, Algorithms, Systems to put Patterns
into Category.
o Relate Perceived Pattern to Previously Perceived
Patterns.
o Learn to distinguish patterns of interest from
their background.

6. hUMan perCeptiOn
o Humans have developed highly sophisticated skills for
sensing their environment and taking actions according to
what they observe, e.g.,
o Recognizing a face.
o Understanding spoken words.
o Reading handwriting.
o Distinguishing fresh food from its smell.
o We would like to give similar capabilities to machines.

7. • Handwritten: sorting letters by postal code.
• Printed texts: reading machines for blind people,
digitalization of text documents.
Optical Character
Recognition
(OCR)
• Face recognition, verification, retrieval.
• Finger prints recognition.
• Speech recognition.
Biometrics
• Medical diagnosis: X-Ray, EKG (ElectroCardioGraph)
analysis.
Diagnostic
systems
• Automated Target Recognition (ATR).
• Image segmentation and analysis (recognition
from aerial or satelite photographs).
Military
applications
ExamplEs of applications

8. Grid by Grid comparison

9. Grid by Grid comparison
A A B
0 0 1 0
0 0 1 0
0 1 1 1
1 0 0 1
1 0 0 1
0 1 1 0
0 1 1 0
0 1 1 0
1 0 0 1
1 0 0 1
No of
Mismatch= 3

10. Grid by Grid comparison
A A BGrid by Grid
Comparison

11. Grid by Grid comparison
A A B
0 0 1 0
0 0 1 0
0 1 1 1
1 0 0 1
1 0 0 1
1 1 1 0
0 1 0 1
0 1 1 1
0 1 0 1
1 1 1 0
No of
Mismatch= 9

12. problEm with Grid by Grid comparison
o Time to recognize a pattern - Proportional to
the number of stored patterns (Too costly
with the increase of number of patterns
stored)
A-Z a-z 0-9
Solution
Artificial
Intelligence
*/-+1@#

13. pattern recognition
o Two Phase : Learning and Detection.
o Time to learn is higher.
• Driving a car
o Difficult to learn but once learnt it becomes
natural.
o Can use AI learning methodologies such as:
• Neural Network.
• Machine Learning.

14. Learning
 How can machine learn the rule from data?
 Supervised learning: a teacher provides a
category label or cost for each pattern in
the training set.
 Unsupervised learning: the system forms
clusters or natural groupings of the input
patterns.

15. o Classification (known categories)
o Clustering (creation of new categories)
cLassification vs. cLustering
Category “A”
Category “B”
Clustering
(Unsupervised Classification)
Classification
(Supervised Classification)

16. pattern recognition process
Post- processing
Classification
Feature Extraction
Segmentation
Sensing
input
Decision

17. pattern recognition process
o Sensing:
• Measurements of physical variables.
• Important issues: bandwidth, resolution , etc.
o Segmentation:
• Removal of noise in data.
• Isolation of patterns of interest from the background.
o Feature extraction:
• Finding a new representation in terms of features.
o Classification
• Using features and learned models to assign a pattern to a
category.
o Post-processing
• Evaluation of confidence in decisions.
Post- processing
Classification
Feature Extraction
Segmentation
Sensing
input
Decision

18. • Collecting training and testing data.
Collect Data
• Domain dependence.
Chose Features.
• Domain dependence.
Chose Model
• Supervised learning
• Unsupervised learning.
Train
• Performance with future data
Evaluate
the Design cycLe

19. Demo
o Online face detector demo:

20. Demo (Cont.)
 With my friend “Albert Einstein”

21. tHAnK
YoU