Unmanned Farmland Treatment with Multispectral Scanning

Unmanned Farmland Treatment with Multispectral Scanning

We use drones equipped with multispectral sensors for precise analysis of vegetation health and soil irrigation.

This makes it possible to:

Detect crop stress before visible symptoms appear;
Identify irrigation deficiencies and water leaks;
Apply fertilizers and herbicides precisely — only where needed;
Create NDVI maps and other vegetation indices;
Analyze crop quality, biomass, and yield potential.

All data is processed using specialized software and delivered to the client as maps, reports, and agronomist recommendations.

Multispectral Sensors Used on Our Drones

Multispectral sensors capture light in several bands — typically in the blue, green, red, red edge, and near-infrared (NIR) ranges. This allows for analyzing plant health, moisture content, and other parameters.

Data Transfer and Processing

We use drones equipped with multispectral sensors for precise analysis of vegetation health and soil irrigation.

Data Collection

The drone with a multispectral sensor flies along a predefined route.
The sensor is synchronized with GPS and IMU — each image contains geotags and metadata.
Data Transfer

Locally, data is stored on an SD card inside the sensor (usually in RAW or TIFF formats).
Wireless transfer (less common) via Wi-Fi / LTE / RTK transmitters — due to data size, offline download is preferred

Practical Example

Scenario:
Goal — detect crop stress and plan variable-rate fertilizer application.
Location — 100 ha field (grain crops).

Equipment:

DJI Matrice 300 RTK — industrial quadcopter with centimeter-level positioning accuracy.
MicaSense RedEdge-MX — multispectral camera with 5 bands (Blue, Green, Red, RedEdge, NIR).

Mission Planning

Task definition by the agronomist — assess crop health before fertilization.
Flight plan in DJI Pilot or DJI Terra software.
Altitude — 100 m (optimal for ~8 cm/pixel GSD).
Image overlap: 75% front, 70% side — for accurate stitching.
Flight speed: 5–6 m/s to avoid motion blur and ensure proper exposure.
Flight time: with ~45 min battery cycles, 3–4 flights required for 100 ha coverage.

Data Collection

The drone flies fully autonomously along the preplanned route.
RTK ensures precise georeferencing of each image.
RedEdge-MX captures images every few meters with set overlap.
Data volume: ~700 multispectral images (5 bands each), ~15 GB total (high-res TIFF).
Radiometric calibration using a reflectance panel before and after flight.

Processing

Data loaded into Pix4Dfields or Pix4Dmapper.

Steps:

Import calibration data.
Stitch multispectral bands.
Build orthomosaic for each spectral band.
Calculate NDVI (optionally NDRE, GNDVI).

Output:

NDVI map in GeoTIFF.
Export to shapefile (SHP) for agronomy software or variable-rate equipment.

Analysis with Agro AI™

Import NDVI, NDRE, and other vegetation maps from Pix4D into Agro AI™.
AI Analysis:
Automatic field classification by crop health zones.
Anomaly detection: diseases, nutrient deficiency, drought stress.
Growth and yield forecasting.
Historical data correlation.

Agro AI™ Output:

Prescription maps for variable-rate fertilizer or pesticide application (ISOXML, SHP, or other equipment-compatible formats).
Tables with dosage rates for each zone.
Recommendations on timing and treatment methods.

Agro AI™ Advantages:

Analysis time reduced from hours to minutes.
Objective results — algorithms trained on thousands of scenarios.
Integration with precision farming systems.

Delivering Results

Formats:
Cloud storage (Google Drive, OneDrive, Dropbox) for instant access.
USB or external SSD if offline.

Delivery includes a mission report with flight parameters and a summary of detected stress zones.

Practical Implementation

Based on Agro AI™ recommendations and NDVI analysis, the agronomist may prescribe:

Variable-rate nitrogen application.
Targeted pesticide treatment.
Adjustments to irrigation schedules.

From flight to decision — Agro AI™ turns data into yield.