C‑DRONE
Agricultural drone flying over a farm field

C-DRONE GUIDE · 1 AUGUST 2026

Crop-emergence counting by drone: plant density, missing plants, replanting decision — price

A few weeks after sowing, only one question matters: how many plants actually emerged, and where? Decisions worth hundreds of euros per hectare hang on the answer — replanting all or part of the field, claiming under the seed guarantee, adjusting fertilisation. Counting on foot means extrapolating from a few sample plots: reliable in the middle of the field, blind to headlands, capped soil and slug patches. The drone counts plant by plant, over the whole area. Here is how an emergence count works, what research says, which crops suit it, and prices observed in 2026.

Published on 1 August 2026, reviewed on 1 August 2026 — regulations in force as of August 2026.

Counting every plant from the sky: what is actually possible

At early stages, each plant stands out against bare soil: a low-altitude flight produces images where 2-4-leaf maize, beet at the spread-cotyledon stage or emerged sunflower can be counted individually, by automatic detection along the rows. The result is not an impression but a figure: plant density per hectare, gap map, deviation from the drill's theoretical seeding rate.

The exercise's reliability is documented: a study by Gnädinger and Schmidhalter published in 2017 in Remote Sensing counted maize plants by drone with detection rates around 95% in favourable conditions, exploiting the contrast between young plants and soil (see the study on Google Scholar). On wheat, where plants are finer and denser, a study by Jin, Liu, Baret, Hemerlé and Comar published in 2017 in Remote Sensing of Environment estimated emergence density from very-low-altitude flights, accurately enough to qualify sowing success (see the study on Google Scholar). The shared constraint: fly at the right stage, neither too early (undetectable plants) nor too late (plants overlapping).

From count to decision: replanting, guarantee, modulation

The figure is only worth the decision it enables. Replanting: in beet or maize, the call is made by comparing actual density with the agronomic threshold below which the lost potential justifies the cost of resowing — the drone map makes it zonal: you resow the waterlogged headland, not the whole 12 hectares. Guarantee and insurance: facing a germination defect, fly, slug or corvid damage, the dated map quantifies the loss and documents the claim with the seed company or insurer. Crop management: heterogeneous density known early lets you modulate nitrogen and calibrate per-zone yield targets.

This early diagnosis dovetails with in-season monitoring — vigour indices, water stress — described in our precision agriculture and water stress by thermal and multispectral drone guides: the emergence count is the first brick, the one that sets the starting potential.

Crops, stages and flight windows

Wide-spaced row crops suit best: maize (2 to 4 leaves), beet (cotyledons to 4 leaves), sunflower (emergence to first leaf pair), oilseed rape at autumn emergence. Wheat and barley are handled as density estimation rather than individual counting, at the very first stages. The window is short — often about ten days per crop — and weather-dependent: it needs drained soil (contrast), even light and little wind.

Practically, the flight is fast (a few minutes per hectare at low altitude), contact-free, and the deliverable arrives within 24 to 72 hours: per-zone density map, overall figure, gap locations, comparison with the theoretical seeding rate if the farmer provides the drill data. Responsiveness rules: a count delivered after the replanting deadline is useless — set that expectation when ordering.

Prices observed in 2026

Ranges observed in 2026 (excl. VAT):

Emergence counting complements the damage assessments (frost, game, slugs) described in our agricultural damage guide, and perennial crops belong to our orchard and fruit-growing guide. Request a quote stating crop, stage, area and decision deadline — spring flight windows book up early.

Frequently asked questions

What accuracy should be expected from a drone emergence count?

On row crops at the right stage, published and observed detection rates commonly sit above 90%, and crucially the error is uniform across the field: the gap map itself is near-exhaustive. That is far better than extrapolating from a few hand-counted sample plots.

Is a multispectral sensor needed to count emergence?

No: at early stages a high-resolution visible sensor is enough in most cases, the plant/soil contrast doing the work. Multispectral becomes useful later in the season, for vigour and stress — a different pass, with different deliverables.

Request a free quote

Put it into practice

Also worth reading