A lightweight Windows application for quantifying fractional green canopy cover from
downward-facing images — built for researchers, extension agents, and educators
who need a reliable, fast, and fully offline tool to analyze hundreds of images.
Fully offline batch processingAnalyze unlimited images of any size. No account, no cloud — your data stays private
Built-in inspectionAdjustable ratio & opacity, and three mask colors to inspect classification before running a full batch
Multiple outputsTabular, binary masks, and blended overlays
Speed toggleProcess at full resolution or ~1 MPx for fast batches
Image metadataFilename, timestamp, GPS coordinates, and device
Noise filtering Removes small isolated specks from the classification
About Canopeo Desktop
Windows · 64-bit
Canopeo Desktop uses the same classification methodology as the original Canopeo App for mobile devices,
validated across multiple crops, growth stages, and lighting conditions in peer-reviewed
studies and more than a decade of field use.
The Canopeo Desktop application is distributed under the PolyForm Noncommercial
license, and developed jointly at Kansas State University and Oklahoma State
University.
Original reference
Patrignani, A., & Ochsner, T. E. (2015). Canopeo: A powerful new tool for
measuring fractional green canopy cover. Agronomy Journal, 107(6), 2312–2320.
doi.org/10.2134/agronj15.0150
Developers A. Patrignani & T. E. OchsnerLicense PolyForm Noncommercial · Contact the developers for commercial inquiries
How Canopeo works
Decision region in chromaticity space
Each panel shows a triangle that contains every color a camera or screen can produce.
Pure red sits at the bottom-right corner, pure green at the top, and pure blue at the
bottom-left. Any color in between — sky, soil, leaves — falls somewhere
inside the triangle. The horizontal and vertical positions give the relative red and
relative green content of a pixel; the blue content is whatever remains.
The four panels show the same set of colors at four different brightness levels, from
dim (V = 0.3) to bright (V = 0.9). The red outline traces the region Canopeo treats
as "green canopy" — the colors a pixel can have for the algorithm to count it as a
leaf rather than soil, residue, sky, or shadow.
Two things stand out. First, the outlined region barely changes shape across the four
brightness levels. That's why Canopeo gives consistent results whether a leaf is photographed
at noon or under overcast skies — the algorithm cares about how green a pixel is compared
to its red and blue components, not about how bright it is overall. Second, the bottom of the
outline shrinks a little at lower brightness. That's the algorithm's safety net for distinguishing
genuinely shadowed canopy (kept) from dark soil with a faint green tint (rejected).
Together, these two behaviors are what make Canopeo robust across crops, growth stages, cameras,
lighting, and field conditions.