The Biosecure Geometry-based System for Hand Modality
From LRDE
The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.
- Authors
- Geoffroy Fouquier, Laurence Likforman, Jérôme Darbon, Bulent Sankur
- Where
- Proceedings of the 32nd IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)
- Place
- Honolulu, Hawaii, USA
- Type
- inproceedings
- Keywords
- Image
- Date
- 2006-12-18
Abstract
We present an identification and authentification system based on hand modality which is part of a reference system for all modalities developed within the Biosecure consortium. It relies on simple geometric features extracted from hand boundary. The different steps of this system are detailed, namely: pre-processing, feature extraction and hand matching. This system has been tested on the Biosecure hand database which consists of 4500 hand images of 750 individuals. Results are detailed with respect to different enrolment conditions such as population size, enrolment size, and image resolution.
Bibtex (lrde.bib)
@InProceedings{ fouquier.07.icassp, author = {Geoffroy Fouquier and Laurence Likforman and J\'er\^ome Darbon and Bulent Sankur}, title = {The Biosecure Geometry-based System for Hand Modality}, booktitle = {Proceedings of the 32nd IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)}, year = 2007, month = apr, address = {Honolulu, Hawaii, USA}, volume = {I}, pages = {801--804}, isbn = {1-4244-0728-1}, abstract = {We present an identification and authentification system based on hand modality which is part of a reference system for all modalities developed within the Biosecure consortium. It relies on simple geometric features extracted from hand boundary. The different steps of this system are detailed, namely: pre-processing, feature extraction and hand matching. This system has been tested on the Biosecure hand database which consists of 4500 hand images of 750 individuals. Results are detailed with respect to different enrolment conditions such as population size, enrolment size, and image resolution.} }