Accession Number:

ADA158030

Title:

Hierarchical Image Segmentation to Infrared Images.

Descriptive Note:

Master's thesis,

Corporate Author:

NAVAL POSTGRADUATE SCHOOL MONTEREY CA

Personal Author(s):

Report Date:

1985-06-01

Pagination or Media Count:

71.0

Abstract:

Recursive image segmentation with hierarchical scope views is a new technique in image processing which systematically divides an image into smaller and smaller quadrants with each region within the quadrants having a structured descriptor. The regions are then processed to remove noise and to check for connected regions across the quadrant boundaries. The quadrants are then brought back together with the obsecured target enhanced and distinguishable from the background. Infrared image data of five different ship targets with the associated noise and interference was processed by this technique with the target information being greatly enhanced. The procedures developed to evaluate the data were found to be inadequate for the task, neccessitating hand evaluation to extract the target. The technique was proven to be a viable solution for extracting target information from infrared data, but very slow processing times and inadequate evaluation procedures limit the usefulness of the program in its present form. Keywords Image segmentation Hierarchical Infrared image Quad split scope view and Recursive segmentation. Theses

Subject Categories:

  • Infrared Detection and Detectors

Distribution Statement:

APPROVED FOR PUBLIC RELEASE