Skip to main content
ARS Home » Research » Publications at this Location » Publication #168036
Title: CALIBRATION OF THRESHOLD LEVELS IN VEGETATION-COVER CLASSIFICATION SOFTWARE

Author
item Booth, D
item Cox, Samuel
item JOHNSON, D - OREGON STATE UNIVERSITY

Submitted to: Society for Range Management Meeting Abstracts
Publication Type: Proceedings
Publication Acceptance Date: 9/1/2003
Publication Date: 1/2/2004
Citation: Booth, D.T., Cox, S.E., Johnson, D.E. 2004. Calibration of threshold levels in vegetation-cover classification software. Abstract. Society for Range Management Meeting Abstracts.

Interpretive Summary:

Technical Abstract: Image analysis software requires users to set a color recognition threshold or class, which is a subjective process that can rsult in human errors and inconsistencies. We developed a manual-grid calibration procedure that objectively accomplishes threshold adjustment. We are measuring the effect of our calibration by comparing digital cover-measurements with on-the-ground point and line intercept sampling methods for study sites on shortgrass and sagebrush steppe, and mountain brush vegetation types. Comparisons made to date using data from 3 sites, imply that image-analysis cover measurements incorporating our calibration method were at least as accurate as measurements derived from conventional procedures.