Invariance of visual perception

Chikhman, V. N. and Shelepin, Yuri and Foreman, Nigel and Passmore, Peter J. and Vakhrameeva, O. and Pronin, Sergey (2008) Invariance of visual perception. Experimental Psychology, 1 . pp. 7-33. ISSN 2072-7593

Full text is not in this repository.

Abstract

Perception and thinking is based largely on the ability of the brain to find in the world invariant features. The purpose of the study – to determine the ranges invariant perception, for various transformations of images of objects (angular size – magnitude, turn corners-projections), under which quantitative characteristics of their perceptions remained unchanged. The recognition psychophysical threshold measuring of incomplete line drawing images (Gollin-test) with different sizes and rotations was provided. The experimental data of invariance perception to the image size transformation and rotating of images are presented. We demonstrate that there is a wide range of angular sizes (from 1.0 up to 50 degrees of visual angle) across which the threshold of recognition of incomplete images does not depend on object size. However, there was found to be a narrow range of small sizes of stimulus (0.19–1, 0 degrees of visual angle) at which there was found to be a dependence of performance on object size. The thresholds of recognition of 3-D objects when viewed as incomplete 2-D images were measured at psychophysical experiments varying perspective of 3-D objects during testing. The memory template of the test objects description has "invariance" properties from 15 up to an angle of 60 degrees. The algorithm for template formation of 3-D objects in the human visual system need to be modeled, taking account of the invariance mechanism. The model of invariant perception, combining well-known model log-polar model and wavelet transform as the primary stage of filtration, and then match filtering and decision-making.

Item Type:Article
Research Areas:Middlesex University Schools and Centres > School of Science and Technology > Computer Science
Middlesex University Schools and Centres > School of Science and Technology > Computer Science > Artificial Intelligence group
Citations on ISI Web of Science:0
ID Code:3363
Useful Links:
Deposited On:04 Dec 2009 06:34
Last Modified:13 Nov 2014 17:30

Repository staff only: item control page

Full text downloads (NB count will be zero if no full text documents are attached to the record)

Downloads per month over the past year