Global to push GA events into
skip to main content

Title: Image processing system and method for recognizing and removing shadows from the image of a monitored scene

The shadow contrast sensitivity of the human vision system is simulated by configuring information obtained from an image sensor so that the information may be evaluated with multiple pixel widths in order to produce a machine vision system able to distinguish between shadow edges and abrupt object edges. A second difference of the image intensity for each line of the image is developed and this second difference is used to screen out high frequency noise contributions from the final edge detection signals. These edge detection signals are constructed from first differences of the image intensity where the screening conditions are satisfied. The positional coincidence of oppositely signed maxima in the first difference signal taken from the right and the second difference signal taken from the left is used to detect the presence of an object edge. Alternatively, the effective number of responding operators (ENRO) may be utilized to determine the presence of object edges.
  1. (Albuquerque, NM)
Issue Date:
OSTI Identifier:
Sandia Corporation (Albuquerque, NM) SNL
Patent Number(s):
US 5495536
Contract Number:
Research Org:
Sandia National Laboratories (SNL), Albuquerque, NM, and Livermore, CA
Country of Publication:
United States
image; processing; method; recognizing; removing; shadows; monitored; scene; shadow; contrast; sensitivity; human; vision; simulated; configuring; information; obtained; sensor; evaluated; multiple; pixel; widths; produce; machine; distinguish; edges; abrupt; difference; intensity; line; developed; screen; frequency; noise; contributions; final; edge; detection; signals; constructed; differences; screening; conditions; satisfied; positional; coincidence; oppositely; signed; maxima; signal; left; detect; presence; alternatively; effective; responding; operators; enro; utilized; determine; difference signal; detection signals; frequency noise; image processing; image sensor; detection signal; /382/