Precision Ag is all about recognizing efficiencies. Whether it’s recognizing optimal irrigation to maximize water usage or ensuring crops get the perfect amount of fertilizer – the success of boosting the overall economics of the farm depend on the accuracy of the information you collect.  

The same is true with imagery – and quite literally, it’s hard to recognize anything if the pictures aren’t clear. Good imagery gives farmers and consultants the ability to accurately detect trouble areas and optimize scouting efforts in a much more targeted approach. 

Yet, strong satellite imagery is not without its challenges, and at the top of that list is cost. “Traditionally it has been difficult for farmers to justify the expense of imagery as an investment that pencils out,” explained Peter Brandt, Chief Technology Officer, Agrian. “As we built out Agrian’s precision toolbox for our user base, we specifically sought to tackle the challenge of expense, as well as other key challenges imagery has presented in the past.”

To check the first box, Agrian offers free imagery to all subscribers. An Agrian user has access to imagery that captures up to two million square miles daily, and provides access to high-resolution, multispectral, in-season imagery for timely extraction of data that directly impact crop production and performance. Those seeking even higher resolution can upgrade to TerrAvion imagery.

Imagery has improved a lot over the past 10 to 15 years, but Agrian continues to find ways to integrate these tools seamlessly and make them more useful to their users. “The more accurate an image is – that is, the more clearly it documents field-specific characteristics and events – the more reliable it is in informing decisions,” said Brandt. “It’s also important that our imagery be customizable, and as easy as possible to read and interpret. That’s what will make it valuable to our advisors and growers.”

Auto-atmospheric correction

Completely unique to Agrian, is their enhanced image processing that provides subscribers a better picture of what’s happening in their fields. The importance of having atmospherically corrected data is found by eliminating the variable of ‘atmosphere’, allowing better comparisons of images. “One way advisors use imagery is to compare the same thing in two different areas,” explained Nick Morrow, VP of Product Management, Agrian. “For example, if I want to look at images from a cornfield in Iowa versus one in Minnesota, potentially collected with a different sensor, in different atmospheres, we can remove that variance in order to better compare.”

Morrow pointed out that correcting images in this way is also helpful when evaluating a single field or area over time. “Across any given month, season or year, the dust and water particles in the air will change; you need to remove those in order to get an accurate representation of what the crop is doing. It’s like the difference between taking a picture of a house across the road on a sunny day compared to a foggy day. If you didn’t know any better, you would think the house had changed, but you are only seeing the effect of the atmosphere.  Remove that, and you can see if the house really changed, and how.”

Agrian offers users a highly advanced method of auto-atmospheric correction called “6S.” Unlike simpler empirical correction methods, 6S uses parameters from the satellite itself to indicate where the sun was positioned at the time of the imaging. This serves as a sort of ‘timestamp’ that can then be used to collect other parameters for that same image, including the scene location and what the atmospheric model should be for that scene at that time of year based on its surface elevation. All of those inputs are used in a model that generates a correction coefficient that is applied to the image, correcting for the effects of the atmosphere between the satellite’s imagery and the surface of the earth.
 
By taking advantage of the platform’s scalable architecture, Agrian developers were able to take an incredibly computationally-intensive step in processing and still deliver quick and accurate imagery data to their users.  

Map Layers allow analysis

The way users interact with imagery has also been made easier. Agrian’s image processing system now eliminates all images of low quality, resulting in the most accurate imaging possible. Map Layers allows users to set up filters for what they want to view – such as date, layer type or imagery type, for quick analysis when making scouting, sampling, seeding prescription, or other agronomic management decisions.  

Brandt believes the Map Layers feature can reveal patterns or insights that might not have been as easy to see before. “It’s not just an analytical tool, but also a discovery tool,” he explained. “Progressive images (or time lapses) are used to create a time series chart that shows how a crop is developing. On that chart, you want to see a smooth curve. If you’re using sub-par images, your crop development curve will be less accurate, more irregular, which makes it harder to interpret in order to make sound management decisions.” 

Zonal statistical tools

For use with the map layers, Agrian also has upgraded its zonal statistical tools, making it faster and easier for users to draw conclusions from customized data sets.

“Users like our tools because it allows them to experiment,” said Morrow. “They enjoy ‘playing’ with map layers, making comparisons, and drawing correlations. We find advisors bringing new insights to growers based on their experience with the statistical tools applied to the images, and we also see growers using the tools to generate questions that they bring to their advisor.  Analyzing imagery really gets conversations going between advisors and growers.”

Speed and accessibility

Although receiving imagery in real-time remains a holy grail, image processing time has been shortened dramatically in recent years. Agrian’s system can consume imagery bands for both Sentinel 2 and Landsat 8 up to six times faster than it had previously. “We get images from satellites and some aerial providers, and within hours – sometimes minutes – make those imagery events available to our subscribers,” said Morrow. 

Cloud computing, centralized processing, greater automation and improved imaging accuracy have all contributed to bringing Agrian users their images so quickly. “It wasn’t long ago that the computational power needed to do that would have made it infeasible. Now, a process that used to take a day or two is done next-day or within hours – and we’re processing more data than ever before.”

Despite all the advancements in technology, Brandt acknowledges that in the real-world, sometimes you need the ability to access your data offline. With Agrian’s native mobile app, images are available offline if users find themselves with spotty or degraded internet service. 

“Imagery has been around a long time,” concludes Brandt, “Our goal has been to build an infrastructure that enables us to retrieve the highest possible volume of imagery at the drop of a hat, and, more importantly, make it quickly available and conveniently useful for our users.” 

Want to learn more about Agrian, or set up a demo to see our software first-hand? Contact us at Sales@agrian.com