Skip to Content
Find More Like This
Return to Search

Fisheye Video Imaging for Diagnosis and Monitoring

Oak Ridge National Laboratory

Contact ORNL About This Technology

Technology Marketing Summary

A variety of investigative settings, from medicine to law enforcement, rely on the clarity
of wide-angle images in order to make accurate interpretations. At ORNL, researchers
developed a computerized video imaging process that maps the coordinates of a
moving image to computer-based reference coordinates. This helps to stabilize and
align images for review.

The ORNL invention addresses several current imaging problems. Medical images
usually have low contrast features. The borders of organs are not sharp and are difficult
to locate using standard imaging techniques. Another problem is that movement
occurs due to respiration, cardiac motion, and blood flow, resulting in poor alignment
through the image sequence and making diagnostic temperature measurements
impossible. Glare caused by lights reflecting off wet tissue surfaces, evaporative
cooling, and random noise also affect clear visualization. Similar challenges exist in law
enforcement, where images from traffic lights are affected by movement and moisture
from adverse weather.

 

Description

Fisheye lens imagery is a wide-angle lens technology for taking wide, hemispherical
images by a mapping technique that evades rectilinear constraints. The ORNL
invention anchors the moving images captured by the fisheye lens during surgery,
making it possible to increase the clarity of organs in areas of low contrast and to
minimize the effects of random motion. This enhances diagnostic accuracy. The
invention registers an image by first segmenting a region and then repeating this
process on multiple images in the same region. From this plurality of segments a
reference image is made. The computer then maps the reference image with the
individual image segments and their subpixel data is then interpolated.

ORNL researchers computed coordinate system maps for each individual image taken
with the reference image. The method uses these coordinates to resample each image,
aligning the reference image directly into the fisheye imaging space. The method also
solves for traffic flow monitoring, vehicle counting, queue detection, and intersection
management.

Benefits
  • Stabilized images are computed directly in the fisheye image space
  • Solves for low organ image contrast in diagnostic medical imaging
  • Solves for motion during surgery, which affects the accuracy of diagnostic analysis of images
  • Stabilizes images for traffic monitoring and intersection management
  • Can be used with other wide-angle optics and/or catadioptric systems


Applications and Industries
  • Potential Applications
  • Medical imaging
  • Law enforcement
  • Traffic
More Information

Patents
Timothy F. Gee and James S. Goddard, Image registration method for medical image sequences, U.S. Patent Application 12/117,478, filed on May 8, 2008.
Timothy F. Gee and James S. Goddard, System and Method for Stabilization of Fisheye Video Imagery, U.S. Patent Application 12/581,484, filed on October 19, 2009.

Lead Inventors
James S. Goddard and Timothy F Gee
Engineering Science and Technology Division
Oak Ridge National Laboratory

Patents and Patent Applications
ID Number
Title and Abstract
Primary Lab
Date
Application 20110091131
Application
20110091131
SYSTEM AND METHOD FOR STABILIZATION OF FISHEYE VIDEO IMAGERY
A coordinate system mapping between each image and the reference image of fisheye images are computed and used to resample each image to align with the reference image directly in the fisheye image space.
10/19/2009
Filed
Patent 8,406,491
Patent
8,406,491
Image registration method for medical image sequences
Image registration of low contrast image sequences is provided. In one aspect, a desired region of an image is automatically segmented and only the desired region is registered. Active contours and adaptive thresholding of intensity or edge information may be used to segment the desired regions. A transform function is defined to register the segmented region, and sub-pixel information may be determined using one or more interpolation methods.
Oak Ridge National Laboratory 03/26/2013
Issued
Technology Status
Technology IDDevelopment StageAvailabilityPublishedLast Updated
1831, 2084DevelopmentAvailable09/25/201209/25/2012

Contact ORNL About This Technology

To: Eugene Cochran<cochraner@ornl.gov>