Best practices to fly and stitch crop imagery
This article provides practical considerations for capturing and processing drone imagery in agricultural settings to ensure the best possible stitching results.
Flexibility is required
Agricultural professionals are skilled at following instructions for a variety of hardware and software. However, when stitching crop imagery, a single set of instructions will not work for all situations. Like many aspects of farming, important decisions about your aerial survey must be made in the field, and you must be flexible.
Experience is key to success. The following are key considerations that can lead to greater success when stitching crop imagery together.
Add unique features to your flight path
Stitching software performs best when it can identify unique inconsistencies, known as tie points, across multiple photos. These unique features can be the difference between a successful stitch and poor results.
Since stitching software picks up on these unique things in your photo pile, ensuring variability is dotted across a field is important.
If your field is bounded by a road, include it in the flight path.
Even if you don't need the corners around a center pivot, capture them anyway, as that variation may help with a successful stitch.
Fly higher for better tie points
Unless you are using DroneDeploy Ag Tools for features like stand count, fly as high as regulations allow to capture the largest area in every shot. More tie points that the stitching software can find between overlapping images increases the likelihood of a usable product.
Flying at 400 feet above ground level (AGL) typically provides Ground Sample Distances (GSDs) of around 4 inches or better in many cases.
Most farm implements operate on scales of feet, not inches, so the resolution from a higher flight is usually sufficient for analysis software such as SMS, SST, or Apex.
Fly efficiently by adjusting the overlap
It is important to maximize the efficacy of your flights by acquiring only the necessary data. Flying higher allows you to cover more ground faster, but you can also save time and battery life by adjusting the image overlap.
DroneDeploy allows you to modify both sidelap (cross-track) and frontlap (along-track).
Frontlap density (shutter frequency) is ultimately determined by the drone’s ground speed. Flying slower consumes more battery on a multi-copter.
Sidelap involves fewer rows, which means less flight time.
Consider a field where both front and sidelap are set to 80%, resulting in a successful stitch. Tweaking the side and frontlap to 65% and 75%, respectively, can provide almost four and a half additional minutes of flight time that can be used elsewhere.
Note: Optimal overlap may change on the same field as the season progresses and crops change. More homogeneous areas (such as a mature, uniform crop canopy) require more overlap to stitch, if they stitch at all.
v2.2