LINEAR SHIFT-VARIANT PROCESSING OF 2-D SIGNALS VIA SEQUENCE CONVOLUTION: SPECIAL CASES AND EXPERIMENTAL RESULTS.

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

A general 2-D linear shift-variant operation can be replaced by a 4-D convolution. If a 'variation band-limited' point spread function is assumed, this 4-D convolution may be carried out as a sequence convolution, i. e. a discrete set of 2-D convolutions, which can be performed very effectively by special convolving networks or coherent optics. The underlying principle of that technique is reviewed. If the point spread function is further restricted to either rotationally or radially symmetrical functions, instead of the 4-D convolution a certain 3-D convolution is sufficient and thus the sequence convolution is drastically simplified. Both special cases are discussed. Experimental results are shown.

Original languageEnglish
Title of host publicationProceedings - International Conference on Pattern Recognition
PublisherIEEE
Pages151-154
Number of pages4
ISBN (Print)0818605456
StatePublished - 1984

Publication series

NameProceedings - International Conference on Pattern Recognition

Fingerprint

Dive into the research topics of 'LINEAR SHIFT-VARIANT PROCESSING OF 2-D SIGNALS VIA SEQUENCE CONVOLUTION: SPECIAL CASES AND EXPERIMENTAL RESULTS.'. Together they form a unique fingerprint.

Cite this