Multimedia Databases (MMDB)

1
Multimedia Databases(MMDB)

• A Content-Based Image Retrieval
  Perspective(CBIR)

2
Types of Media Files

• Static media images and handwriting
• Dynamic media video and sound bytes
• Dimensional media 3D games or CAD)

3
MMDB Motivation Factors

• Acquisition email, phone, web sites like FLIKR
• Generation camera phones, digital cameras,
• Storage databases design
• Processing power and techniques more
  sophisticated
• Huge increase in multimedia data on computers and
  their transmission over networks.

4
Database Support

• Databases provide consistency, concurrency,
  integrity, security and availability of data for
  the large amount of multimedia data available.
• From a user perspective, databases provide
  functionalities for manipulation and querying the
  huge collections of stored data.

5
Media Data Stored

• Media data actual media data representing
  images, audio, and video that are captured,
  digitized, processed, compressed and stored.
• Media format data consists of media data format
  stored after the acquisition, processing, and
  encoding phases. Examples are sampling rate,
  resolution, frame rate, encoding scheme etc.
• Media keyword data For example, for a video this
  might include the date, time, and place of
  recording , the person who recorded, the scene
  that is recorded. Also called content descriptive
  data.
• Media feature data This contains the features
  derived from the media data. A feature
  characterizes the media contents. For example,
  this could contain information about the
  distribution of colors, the kinds of textures and
  the different shapes present in an image. This is
  also referred to as content dependent data.

6
Image Retrieval (IR)

• Image Retrieval is the process of searching and
  retrieving desired images from a large database.
• IR provides resourceful use of prolific image
  data
• The efficiency of implementations have increased
  over the past two decades

7
IR Methodology

• A simple image retrieval implementation uses
  individually entered keywords or descriptions of
  inserted images so that retrieval is performed
  over the annotations in normal textual forms. 
• If an image is poorly or incorrectly annotated,
  or a poor choice of arbitrary query values are
  given by the user then the desired output is not
  received even if it exists in the database.
• Therefore, a lot of research has gone into
  automatic annotation of image description.

8
Image Features Stored

• Color Red, Blue, Green, etc
• Texture Similarity in grouping of pixels
• Shape Edge detection
• Spatial Spacing of Features
• Semantic Correlated description of image data.
  E.G Color blue, Shape Large, Texture
  smooth, Spatial Top of image Sky

9
Feature Extraction

• Image Segmentation open ended topic
• Segment Classification based off characteristics
• Filtering Techniques Extract image features such
  as texture by passing images through a filter

10
Feature Application In DB

• Users could supply a range for color, texture, or
  shape for queries
• Features can be generated on a typical semantic
  set for automatic annotation of new pictures

11
Content based image Retreival (CBIR)

• Avoids the necessary use of textual descriptions
• Organizes digital archives by visual content
• Retrieves images based on visual similarity to a
  user-supplied query image or image features.

12
Query Types

• Keyword common text searching techniques
• Feature Ex. Draw area for location and size.
  Select color regions. Select shape. B-tree is
  traversed based off given index value.
• Semantic Provide words to describe feature sets
  that are used to query a database
• Composite Index involves combination of above

13
Query examples from CIBR at the end of the early
years
14
Content-Based

• Straight-forward implementation is each feature
  is used as an index. Not very efficient for
  querying
• Create an index as described earlier as a
  combination of region classification, spatial
  location, shape, and color. EX. 20-bit index
  key 3bits location, 8 bits color, 4 bits size, 5
  bits shape. B-tree indexing method is used.

15
Relevance Feedback

• A query modification technique attempts to
  improve retrieval performance through iterative
  feedback and query refinement.
• Used in ALIPR

16
Data Flow From CBIR at the end of early years
17
IR Implementation Examples

• Yahoo or Google Image Searches based mainly on
  annotated description and filename
• Automatic Linguistic Indexing of Pictures
  (ALIPR) learning algorithm that annotates with
  feedback

18
Future Work

• The open ended nature of image segmentation
  restricts the accuracy of object recognition. As
  segmenters improve so will the databases
  capability.
• The integration of image retrieval can be
  implemented in computer vision applications.
• Many researchers believer that image retrieval
  has grown out of its infancy and now focus will
  be on applications and proliferating algorithms
  into indivduals lives.

19
Bibliography

• Datta, R., Joshi, D., Li, J., and Wang, J. Z.
  2008. Image retrieval Ideas, influences, and
  trends of the new age. ACM Comput. Surv. 40, 2
  (Apr. 2008), 1-60. DOI http//doi.acm.org/10.1145
  /1348246.1348248
• Stanchev P., Using Image Mining for Image
  Retrieval, IASTED International Conference
  Computer Science and Technology, May 19-21,
  2003, Cancun, Mexico.
• "Multimedia Database." Information Technology
  Portal (IT Portal) - India. Web. 04 Dec. 2009.
  lthttp//www.peterindia.net/MultimediaDatabase.html
  gt.
• "Image retrieval -." Wikipedia, the free
  encyclopedia. Web. 04 Dec. 2009.
  lthttp//en.wikipedia.org/wiki/Image_retrievalgt.
• Smeulders, Content-based image retrieval at the
  end of the early years A.W.M.JournalIEEE
  transactions on pattern analysis and machine
  intelligence, 2000, Vol22, 12, 1349