Skip to content

Green Area Index with PlantCV 🟢📐

Green Area Index (GAI) tells you how much green canopy is covering your growing area. It rises as your microgreens grow, so measuring it over several days gives you a growth curve — a real, quantitative result you can compare between species, light levels, or watering treatments.

This protocol turns a single top-down photo of a tray into a GAI number using PlantCV (free, open-source plant computer vision) and the AstroCalibration marker (the AstroBotany Spectrum sticker) for size and colour calibration.

What GAI means here

From one top-down photo you measure the projected green area ÷ the ground (tray) area. This is really canopy cover (0 = bare soil, 1 = fully covered). True GAI can exceed 1 when leaves overlap, which a single photo can't see — but for the classroom, canopy cover is an excellent, honest proxy that tracks biomass well. Take photos at the same height and angle each day so they stay comparable.

What you'll need

The idea, in plain language

  1. Colour-correct the photo using the marker, so greens look consistent every day.
  2. Find the green pixels (the canopy).
  3. Convert pixels → cm² using the marker's known size.
  4. Divide green area by the tray's ground area to get GAI.

Step-by-step

import numpy as np
from plantcv import plantcv as pcv

# 1. Read the top-down tray photo
img, path, name = pcv.readimage(filename="tray_day8.jpg")

# 2. Colour-correct using the AstroCalibration marker.
#    auto_correct_color finds the colour card automatically.
#    Set color_chip_size to your sticker's chip size in mm, e.g. (10, 10).
img_cc = pcv.transform.auto_correct_color(rgb_img=img, color_chip_size=(10, 10))

# 3. Set the scale (pixels per cm). The robust classroom method: measure your
#    marker's known length in pixels once (e.g. in PlantCV or ImageJ) and set it here.
px_per_cm = 60.0          # <-- calibrate this from YOUR marker

# 4. Segment the green canopy. The LAB 'a' channel separates green plant tissue
#    from soil and background; green tissue is "dark" in this channel.
a = pcv.rgb2gray_lab(rgb_img=img_cc, channel="a")
green = pcv.threshold.binary(gray_img=a, threshold=120, object_type="dark")
green = pcv.fill(bin_img=green, size=50)     # remove small specks of noise

# 5. Restrict measurement to the growing area of the tray.
roi = pcv.roi.rectangle(img=img_cc, x=200, y=150, w=1200, h=900)
green = pcv.roi.filter(mask=green, roi=roi, roi_type="partial")

# 6. Measure and compute Green Area Index.
green_cm2  = np.count_nonzero(green) / (px_per_cm ** 2)
ground_cm2 = (1200 * 900) / (px_per_cm ** 2)        # the ROI area in cm^2
gai = green_cm2 / ground_cm2                          # 0 = bare, ~1 = full cover

print(f"Green area = {green_cm2:.1f} cm^2 | Ground = {ground_cm2:.1f} cm^2 | GAI = {gai:.2f}")

Tune two numbers for your setup

The threshold=120 and the ROI box (x, y, w, h) are examples. Display the a channel and the green mask while you adjust them so the mask covers the plants and nothing else. Once they work for your camera setup, keep them the same for every photo in a series.

Reading your results

  • GAI near 0 → mostly bare growing media (early days).
  • GAI rising over time → your canopy is filling in. Plot GAI vs. day to see the growth curve.
  • GAI plateauing near 1 → the tray is fully covered (from the top); switch to fresh weight/biomass to keep measuring growth after this point.

Make it a real experiment

  • Time series: photograph the same tray on days 4, 8, and 12 and plot GAI vs. day.
  • Compare: run two species (or two light levels) side by side and compare their GAI curves.
  • Share: record your GAI values, the marker size, px_per_cm, and threshold in your science journal so others can reproduce your analysis (this is what makes it FAIR).

Going deeper

Exact function arguments and output names can change between PlantCV versions — check the docs above for the version you installed. The next stage applies the same calibrated imaging to roots: Root System Architecture with PlantCV.