Accession Number:

ADA090245

Title:

Threshold Selection Using Quadtrees

Descriptive Note:

Technical rept.

Corporate Author:

MARYLAND UNIV COLLEGE PARK COMPUTER VISION LAB

Personal Author(s):

• Wu, Angela Y.
• Hong, Tsai-Hong
• Rosenfeld, Azriel

Report Date:

1980-03-01

Pagination or Media Count:

19.0

Abstract:

The quadtree representation of binary arrays can be generalized to a quadtree approximation for images. A block is subdivided if its gray level standard deviation is greater than e.g. half of the global standard deviation. As a result, small blocks will tend to occur near region borders, and suppressing these blocks deepens the valleys on the images histogram, thus making it easier to select a threshold for extracting objects from their background in the image.

Descriptors:

• *IMAGE PROCESSING
• *PATTERN RECOGNITION
• BINARY NOTATION
• GRAPHS
• GRAY SCALE
• HISTOGRAMS
• MEAN
• NODES
• SEGMENTED
• STANDARD DEVIATION
• THRESHOLD EFFECTS

Subject Categories:

• Human Factors Engineering and Man Machine Systems

Distribution Statement:

APPROVED FOR PUBLIC RELEASE