Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Object detection in natural backgrounds predicted by discrimination performance and models

A M Rohaly1, A J Ahumada, A B Watson

  • 1U.S. Army Research Laboratory, Aberdeen Proving Ground, MD 21005, USA. amrohaly@arl.mil

Vision Research
|January 13, 1998
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Partial gastrectomy for simple ulcer; a review of the end-results of 132 cases, with a criticism of the Polya operation.

The British journal of surgery·2010
Same author

A cross-sectional study of psychological morbidity in patients with acne, psoriasis and atopic dermatitis in specialist dermatology and general practices.

Journal of the European Academy of Dermatology and Venereology : JEADV·2008
Same author

Discriminability measures for predicting readability of text on textured backgrounds.

Optics express·2002
Same author

Visual detection of spatial contrast patterns: evaluation of five simple models.

Optics express·2002
Same author

Mondor's disease of the breast: sonographic and mammographic findings.

AJR. American journal of roentgenology·2001
Same author

Allergic contact dermatitis due to benzalkonium chloride in plaster of Paris.

The Australasian journal of dermatology·2001

Image discrimination models can predict object visibility in natural scenes. A multiple channel model with a contrast gain factor and a summation exponent of 4 showed the best performance, closely matching human observer data.

Area of Science:

  • Computer Vision
  • Human Visual Perception
  • Image Processing

Background:

  • Image discriminability models aim to predict the visibility of differences between images.
  • Object detection in natural scenes remains a challenge for computational models.

Purpose of the Study:

  • To compare the predictive accuracy of three image discrimination models for object detectability in natural backgrounds.
  • To evaluate the impact of different summation exponents and a contrast gain factor on model performance.

Main Methods:

  • Three models were tested: a multiple channel Cortex transform model, a single channel contrast sensitivity filter model, and a digital image difference metric.
  • Each model employed a Minkowski distance metric with summation exponents of 2, 4, and infinity.
  • Model predictions were compared against object detectability data from 19 human observers.
Keywords:
NASA Center ARCNASA Discipline Space Human Factors

Related Experiment Videos

Main Results:

  • Without a contrast gain factor, the multiple channel model with a summation exponent of 4 performed best (RMS error of 2.3 dB).
  • A contrast gain factor significantly improved predictions for all models and exponents.
  • With the contrast gain factor, all models achieved near 1 dB prediction error, with exponent 4 being optimal across models.

Conclusions:

  • Image discrimination models, particularly the multiple channel model with appropriate parameters, can effectively predict object detectability in natural scenes.
  • The inclusion of a contrast gain factor enhances the predictive power of these models.
  • These findings support the utility of computational models in understanding visual performance in complex environments.