Source codes of Locally Likely Arrangement Hashing (LLAH) written by Tomohiro Nakai

This repository contains the source codes of Locally Likely Arrangement Hashing (LLAH), a real-time camera-captured document image retrieval method, written by Tomohiro Nakai.

Related papers are found in the publication database of Intelligent Media Processing Laboratory, Department of Computer Science and Intelligent Systems Graduate School of Engineering, Osaka Prefecture University.

Old stable LLAH source code

Directory: ./old_stable Visual Studio Project: ./vs/llahdoc080620/

This directory contains an old version of LLAH program, which is fast and stable.

Usage:

• Database construction: llahdoc -c [document image directory] [document image suffix] [option]
• Retrieval: llahdoc [option] [query image]
• Real-time retrieval: llahdoc -S [option] (server.ini is needed to be setup)

See Function AnalyzeArgAndSetExtern2 in init.c for the options.

Representative references:

• Tomohiro Nakai, Koichi Kise, Masakazu Iwamura.
  Use of Affine Invariants in Locally Likely Arrangement Hashing for Camera-Based Document Image Retrieval.
  Lecture Notes in Computer Science (7th International Workshop DAS2006), 3872, pp.541-552, Nelson, New Zealand (2006-2).

• Tomohiro Nakai, Koichi Kise, Masakazu Iwamura.
  Camera Based Document Image Retrieval with More Time and Memory Efficient LLAH.
  Proc. Second International Workshop on Camera-Based Document Analysis and Recognition (CBDAR2007), pp.21-28 (2007-9).

• Tomohiro Nakai, Koichi Kise, Masakazu Iwamura.
  Real-Time Document Image Retrieval with More Time and Memory Efficient LLAH.
  Proc. Second International Workshop on Camera-Based Document Analysis and Recognition (CBDAR2007), pp.168-169 (2007-9).

• Masakazu Iwamura, Tomohiro Nakai, and Koichi Kise.
  Improvement of Retrieval Speed and Required Amount of Memory for Geometric Hashing by Combining Local Invariants.
  Proc. 18th British Machine Vision Conference (BMVC2007), 2, pp.1010-1019 (2007-9).

• Tomohiro Nakai, Koichi Kise, Masakazu Iwamura.
  Real-Time Retrieval for Images of Documents in Various Languages.
  Handout of 8th IAPR International Workshop on Document Analysis Systems (DAS2008), p.37 (2008-9).

• Tomohiro Nakai, Koichi Kise, Masakazu Iwamura.
  Real-Time Retrieval for Images of Documents in Various Languages using a Web Camera.
  Proceedings of the 10th International Conference on Document Analysis and Recognition (ICDAR2009), pp.146-150, Barcelona, Spain (2009-7).

Annotation extraction

Directory: ./anno_extract Visual Studio Project: ./vs/annex

Usage: annex [option] [original image] [annotated image]

See function InitAnnex in annex.c for the options.

Reference:

• Tomohiro Nakai, Koichi Kise, Masakazu Iwamura.
  A Method of Annotation Extraction from Paper Documents Using Alignment Based on Local Arrangements of Feature Points.
  Proc. 9th International Conference on Document Analysis and Recognition (ICDAR2007), 1, pp.23-27 (2007-9).

Mosaicing

Directory: ./llahmos Visual Studio Project: ./vs/llahmos

Usage: llahmos [option] [image1] [image2]

See function AnalyzeArgAndSetExtern2 in annex.c for the options.

Reference:

• Tomohiro Nakai, Koichi Kise, and Masakazu Iwamura.
  Camera-based document image mosaicing using LLAH.
  Proc. Document Recognition and Retrieval XVI (Proc. SPIE Volume 7247), 7247-25 (2009-1).

Demo server

Directory: ./demo_server Visual Studio Project: ./vs/llahdoc

This directory contains the server program for the real-time demonstration.

Usage:

• Database construction: llahdoc -c [document image directory] [document image suffix] [option]
• Retrieval: llahdoc [option] [query image]
• Real-time retrieval: llahdoc -S [option] (server.ini is needed to be setup)

See function AnalyzeArgAndSetExtern2 in init.c for the options.

Demo client

Visual Studio Project: ./vs/dcamc

Usage:

1. Launch the server program
2. Plug in a web camera
3. Run the client program

client.ini is needed to be setup (In particularly, the name of client should be correctly written).