Data Search Tool For Face Recognition

1
Data Search Tool For Face Recognition
COMP6703 Demonstration
Supervisor Professor Tom Gedeon

• Presented by
• Student Yan Gao
• Student Id u2581609

2
Overview

• What is aim for this project???
• The aim of this project is to find
• often occurring conditions in a
• range of face images.

3
Requirement

• What is requirement for this project???
• The software is designed as a tool with
• user interface, for face recognition data
• search. The tool allows users to set the
• search condition, and the software is able
• to search rapidly through a large number
• of images in local disk, and to keep
• records of any occurrence of a defined
• condition.

4
Center-Surround Processing

• Why we use Center-Surround processing???
• Centre-surround is a simple way of
• preprocessing an image, which is like
• finding lines. Basically if a pixel has
• mostly white around it, it is an empty
• area, if its half white and half black, then
• its probably an edge, etc.

5
Center-Surround Processing Example

• Example

6
Key Terminology (1)

• Local Search --- In the process of local search,
  users are able to do a search to find all matched
  pixels in different position within the defined
  image.
• Normal Search --- It is possible for user to
  search through face images with the same shot
  angle on the local disk, and find matched pixels
  at the exact same position with the defined
  condition. In the program, it is one types of
  same location search.

7
Key Terminology (2)

• Rough search --- It is another type of same
  location search, the only difference between
  normal search and rough search is that it is
  possible for user to search matched pixels with 0
  to 5 pixels variation.
• Variable search --- it means that user can define
  search type and the range of search images
  through parameters in the user interface, then
  start a search to find the matched pixels with
  the defined condition.

8
File System Structure

• File System Structure

9
Development Process

• Five Steps in Development
• Requirement Analysis
• Modeling
• Implementation
• Testing and Debugging
• Deployment and Maintenance

10
User Cases (1)

• User starts the program.
• User starts a local search.
• User starts a normal search.
• User starts a rough search.
• User starts a variable search.
• User sets the search condition as one or more
  single pixel.
• User sets the search condition as pixels domain
  in rectangle or square shape.

11
User Cases (2)

• User sets the search condition as pixels domain
  in ellipse or round shape.
• User chooses the rough search pixels, and sets
  the variation parameter.
• User saves the search result as a dat file.
• User imports the record dat file, and displays it
  in image mode.
• User imports the record dat file, and displays it
  in text mode.

12
Domain Chart

• Domain Chart

13
Gui Domain (1)

• Mission Statement
• This domain provides a graphical
• user interface for users to access
• functions of the software.

14
Gui Domain (2)

• Class Diagram

15
Conditionsetting Domain(1)

• Mission Statement
• This domain is responsible for setting
• search condition. There are two sorts of
• search condition one is defined as one or
• more single pixel, another is defined as a
• rectangle or round shape pixel domain.

16
Conditionsetting Domain(2)

• Class Diagram

getCs() getPy() getPx() getValidFlag()
countSurround () addPixSingle () mouseClicked ()
grabPixels()
17
DataImplementation Domain (1)

• Mission Statement
• This domain is responsible for image
• data search, and the path of matched
• Images are displayed in a list of
• ResultTreePanel class, and the matched
• pixels are shown in the ResultImagePanel
• class.

18
DataImplementation Domain (2)

• Class Diagram

LocalSearch () RoughSearch () NormalSearch ()
VariableSearch ()
getMessage () mouseClicked () paintComponent ()
displaySLSearch () displayLSearch ()
displayVSearch ()
19
RecordDisplaying Domain

• Mission Statement
• This domain is used to display some
• information reading from dat file.

20
Next Step

• An Example About
• Variable Search

21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24

• THANKS
• FOR YOUR LISTENNING
• Any Questions???