Abstract
We designed and implemented a composite-wavelet-matched filter that
is invariant to continuous-scale changes of the input and is useful for
correlation-based pattern recognition of objects whose size is not
known exactly. We optimize the adaptive wavelet to extract sparse
image features and use the scale–space analysis to determine the
wavelet scale for the scale invariance. Experimental results
obtained by use of a programmable optical correlator with three
liquid-crystal spatial light modulators are demonstrated.
© 1999 Optical Society of America
Full Article |
PDF Article
References
You do not have subscription access to this journal. Citation lists with outbound citation links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.
Contact your librarian or system administrator
or
Login to access OSA Member Subscription
Cited By
You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.
Contact your librarian or system administrator
or
Login to access OSA Member Subscription
Equations (16)
You do not have subscription access to this journal. Equations are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.
Contact your librarian or system administrator
or
Login to access OSA Member Subscription
Metrics
You do not have subscription access to this journal. Article level metrics are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.
Contact your librarian or system administrator
or
Login to access OSA Member Subscription
Share with Facebook
Tweet This
Post on reddit
Share with LinkedIn
Add to CiteULike
Add to Mendeley
Add to BibSonomy