Title: Gamera A Toolkit for Structured Document Recognition including Music
1GameraA Toolkit for Structured Document
Recognition including Music
- Ichiro Fujinaga
- McGill University
- with
- Michael Droettboom, Karl MacMillan, G. Sayeed
Choudhury, Tim DiLauro, Mark Patton, Teal
Anderson - Levy Project II
- Digital Knowledge Center
- Sheridan Libraries
- Johns Hopkins University
2Content
- Levy Project
- Levy Sheet Music Collection
- Digital Workflow Management
- Optical Music Recognition
- Gamera
- Guido / NoteAbility
3Lester S. Levy Collection
4Lester S. Levy Collection
- North American sheet music (17801960)
- Digitized 29,000 pieces
- including The Star-Spangle Banner and Yankee
Doodle - Database of
- text index records
- images of music (8bit gray)
- lyrics (first lines of verse and chorus)
- color images of cover sheets (32bit)http//levysh
eetmusic.mse.jhu.edu
5Digital Workflow Management
- Reduce the manual intervention for large-scale
digitization projects - Creation of data repository (text, image, sound)
- Optical Music Recognition (OMR)
- Gamera
- XML-based metadata
- composer, lyricist, arranger, performer, artist,
engraver, lithographer, dedicatee, and publisher - cross-references for various forms of names,
pseudonyms - authoritative versions of names and subject terms
- Music and lyric search engines
- Analysis toolkit
6Optical Music Recognition (OMR)
- Trainable open-source OMR system in development
since 1984 - Staff recognition and removal
- Lyric removal
- Stems and notehead removal
- Music symbol classifier
- Score reconstruction
- Lyric classifier?
7The problem
- Suitable OCR for lyrics not found
- Commercial OCR systems are often inadequate for
non-standard documents - The market for specialized recognition of
historical documents is very small - Researchers performing document recognition often
re-invent the basic image processing wheel
8The solution
- Provide easy to use tools to allow domain experts
(people with specialized knowledge of a
collection) to create custom recognition
applications - Generalize OMR for structured documents
9Introducing Gamera
- Framework for creation of structured document
recognition system - Designed for domain experts
- Image processing tools (filters, binarizations,
etc.) - Document segmentation and analysis
- Symbol segmentation and classification
- Feature extraction and selection
- Classifier selection and combiners
- Syntactical and semantic analysis
- Generalized Algorithms and Methods for
Enhancement and Restoration of Archives
10Features of Gamera
- Portability (Unix, Windows, Mac)
- Extensibility (Python and C plugins)
- Easy-to-use (experts and programmers)
- Open source
- Graphic User Interface
- Interactive / Batchable (scripts)
11Architecture of Gamera
Graphic User Interface (wxWindows)
Plugins (Python)
Plugins (C)
GAMERA Core (C)
12Example of C Plugin
- // Number of pixels in matrix
- include gamera.hh
- ifdef __area_wrap__
- define NARGS 1
- define ARG1_ONEBIT
- endif
- using namespace Gamera
- template ltclass Tgt
- feature_t area(T m)
- return feature_t(m.nrows() m.ncols())
13Example of Python Plugin
- // This filters a list of CC objects
- import gamera
- def filter_wide(ccs, max_width)
- tmp
- for x in ccs
- if x.ncols() gt max_width
- x.fill_matrix(0)
- else
- tmp.append(x)
- return tmp
14Gamera Interface(screenshot in Linux)
15Gamera Interface(screenshot in Linux)
16Histogram(screenshot in Linux)
17Thresholding(screenshot in Linux)
18Thresholding(screenshot in Linux)
19Staff removal Lute tablature
20(No Transcript)
21Classifier Lute(screenshot in Linux)
22Staff removal Neums
23Classifier Neums(screenshot in Linux)
24Greek example
25GUIDO Music Notation FormatH. Hoos, K. Renz, J.
Kilian
- A formal language for score-level
representation - Plain text readable, platform independent
- Extensible and flexible
- Adequate representation
- NoteServer Web/Windows
- GUIDO/XML
- NoteAbility (K. Hamel)
26GUIDO An example
\beamsOff \cleflt"treble"gt \keylt"D"gt
f1/8. g1/16 a1/4. d21/8 d1/4. c1/8
e11/2 _1/4 f1/8. g1/16 c21/4. b11/8
a1/4. g1/8 e1/2 f1/4 f1/8. g1/16
a1/4. d21/8 d1/4. c1/8 e11/2 _1/4
f1/8 g c21/4. b11/8 a1/4. c1/8 ,
27NoteAbility Demo
28Conclusions
- Gamera allows rapid development of
domain-specific document recognition applications - Domain experts can customize and control all
aspects of the recognition process - Includes an easy-to-use interactive environment
for experimentation - Beta version available on Linux
- OS X version in preparation
29Acknowledgements
- National Science Foundation
- Institute of Museum and Library Services
- The Levy Family
30OMR Classifier
- Connected-component analysis
- Feature extraction, e.g
- Width, height, aspect ratio
- Number of holes
- Central moments
- k-nearest neighbor classifier
- Genetic algorithm
31Overall Architecture for OMR
Image File
Staff removal Segmentation
Recognition K-NN Classifier
Output Symbol Name
Optimization Genetic Algorithm K-nn Classifier
Knowledge Base Feature Vectors
Best Weight Vector
Off-line