Logo image
A solution to the feature correspondence problem inspired by visual scanpaths
Journal article   Peer reviewed

A solution to the feature correspondence problem inspired by visual scanpaths

Jeffrey L Krichmar, Kim T Blackwell, Garth S Barbour, Alexander B Golovan and Thomas P Vogl
Neurocomputing (Amsterdam), Vol.26, pp.769-778
06/01/1999
DOI: 10.1016/S0925-2312(98)00146-5

View Online

Abstract

The organization of features into ordered sets to allow comparison of visual images is called the correspondence problem. When presented with a familiar image, humans scan an image in a scene with a specific and distinctive pattern. We suggest that the visual system uses scanpaths to solve the feature correspondence problem. A computer algorithm that simulates this process was developed and tested on different classes of images. In agreement with psychophysical data, the algorithm processed familiar images with idiosyncratic “eye movement” patterns and with fewer “eye movements” than unfamiliar patterns; additional fixation time was spent on unexpected patterns. This solution of the correspondence problem is well suited to computer vision systems.
Computer vision Eye movements Feature correspondence Neural networks Scanpaths

Details

Metrics

15 Record Views
Logo image