Spatial Scale in Image Detection and Recognition.

Abstract

Considerable recent physiological and psychophysical evidence suggests that the visual system operates as a series of independent channels or analyzers, each sensitive to image structure at a different spatial scale. In this view, image structure is processed separately at different scales by the various channels. Several individuals have argued that the broad, low-frequency channels respond to global or Gestalt properties of an image and are important in early processing--for instance, during an initial glance at an image. In constrast, the high-frequency channels are sensitive to local detail and are important in later visual processing when attention has been focused on a particular aspect of the image. Two experiments investigated the ability of human observers to detect and recognize simple objects in visual images. Prior to presentation, the images were transformed by spatial frequency filters to emphasize the global- (low spatial frequencies), local- (high spatial frequencies) or intermediate- (mid spatial frequencies) scale structure. Four categories of top-view ship hulls were synthesized for the experiments. In the first experiment separate groups of observers made both detection (which quadrant of the display contained a ship?) and recognition (which of the four ships occurred?) judgements. In the second experiment , observers also selected the filter condition to be display on each trial prior to the detection or recognition response.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Feb 01, 1986
Accession Number
ADA171348

Entities

People

  • James H. Howard Jr.

Organizations

  • The Catholic University of America

Tags

Communities of Interest

  • Biomedical
  • Ground and Sea Platforms
  • Human Systems
  • Space
  • Weapons Technologies

DTIC Thesaurus Topics

  • Air Force
  • Applied Psychology
  • Artificial Intelligence
  • Biological Sciences
  • Computer Science
  • Computer Vision
  • Detection
  • Detectors
  • Frequency Bands
  • Human Factors Engineering
  • Image Recognition
  • Information Processing
  • Information Science
  • Military Research
  • Psychology
  • Two Dimensional
  • Visual Perception

Readers

  • Atmospheric Science / Meteorology, specifically Wind Wave Turbulence.
  • Computer Vision.
  • Vision Science/Vision Psychology/Cognitive Neuroscience.