First, you’ll need to create a survey flight pattern for your drone.
The drone takes off and flies this survey pattern. A mounted camera takes photos the entire time.
There are a few critical requirements for any survey:
A single survey can consist of a single flight or multiple consecutive flights over a contiguous area.
For a single-flight survey…
If your photos are automatically geotagged in flight, simply put all the relevant survey photos into a zip file before uploading to FarmLens™.
If you are using FarmLens Uploader™ to assist with geotagging, the time range of your photo set must fall within the time range of the flight log (recorded by your ground control device). To ensure this, always conform to this sequence: Before the survey, 1) start recording the flight log and 2) initiate camera triggering; After the survey 1) stop camera triggering and 2) end the flight log recording. This way, FarmLens Uploader™ will always be able to find the correct flight log to geotag your photo set.
For a multi-flight survey…
If your photos are automatically geotagged in flight, include the relevant survey photos from each individual flight in a single zip file before uploading to FarmLens™.
If you are using FarmLens Uploader™, rather than creating separate photo sets and flight logs for each flight, create a single photo set and flight log for the entire series of flights. Find details on how to do this here.
Must have GPS
Must be able to carry a survey-capable camera mounted so the camera is orthogonal to the ground.
Some tried and true survey cameras: Modified GoPro Hero 4, CanonS100, Canon S110/S120.
1) In-flight geotagging by camera’s GPS or by drone’s GPS.
or
2) Drone with a 3D Robotics flight controller for post-flight geotagging w/ FarmLens Uploader™.
(All parameters based upon Agribotix’s own hardware specs)
• Best if overhead, bright, and uniform.
• If there is cloud cover, it must be consistent and provide uniform light filtration.
• Cloud shadows reduce NIR reflectance and can invalidate this as a measure of crop health.
• Surveys done early on the morning or late in the day may not work for anything more than regular RGB imagery. The low sun angle will cast shadows from trees or relatively higher crop areas. This can invalidate NIR reflectance as a measure of crop health.
• Too early or late in the day and the sun angle is so low that it casts shadows.
• At midday, there is the possibility of solar reflectance. This can happen with crops that have large, shiny leaves or fields that are flooded or soggy. The field becomes a mirror with an especially bright spot from the sun in each photo. In the final orthomosaic, this can create spots or stripes that are not indicative of crop health.
• No precipitation
• No/minimal wind
• No cloud cover is best. If there is cloud cover, it must be consistent and provide uniform light filtration.
Wind speed: under 20 mph (32 kph)
Ground speed: 14 m/s
Altitude: 100 m
Pattern: Lawnmower patter. ‘Survey (Grid)’ in Mission Planner.
Survey Grid Angle: if windy, perpendicular to wind direction. If no wind, oriented for least switchbacks.
Photo Interval (if not controlled by autopilot): 2 seconds
Photo Sidelap: 80%
Photo Overlap: 80% (achieved w/ camera interval of 2 seconds when traveling at 14m/s and 100m altitude)
Survey Grid Area: 20 meters beyond edge of the target survey area
See this blog post for additional details.
All parameters same as Standard except…
Photo Interval (if not controlled by autopilot): increase camera interval rate to 1 second. If experiencing stitching failure, increase rate to 1/2 second.
See this blog post for additional details.
Wind Speed: Under 20 mph (32kph)
Ground Speed: 8 m/s
Altitude: 20m
Pattern: Lawnmower pattern. ‘Survey (Grid)’ in Mission Planner.
Survey Grid Angle:
If flying with a gimbal:
• If light or no wind, fly perpendicular to the direction of rows (optimize count accuracy).
• If strong wind, fly perpendicular to wind direction (optimize flight efficiency).
If flying with NO gimbal:
• If no wind, fly perpendicular to the direction of rows (optimize count accuracy).
• If any wind, fly perpendicular to wind direction (maintain consistent camera angle and optimize flight efficiency).
Camera ISO: 400
Photo Interval (if not controlled by autopilot): 2 seconds
Photo Sidelap: 10%
Photo Overlap: 10%
Survey Grid Area: 10 meters in from edge of field. DO NOT include the perimeter of the field in the survey. Only survey the planted area of the field. Inclusion of the perimeter and area beyond may invalidate survey results.