Agribotix images: An X-ray for your field

A lot of buzz surrounds the use of UAVs for precision agriculture, but what can you as a farmer or agronomist actually do with the images? The Agribotix Hornet can take two different types of images of your field: color and infrared. As we wrote about in a previous blog post, the infrared images can be further processed into vegetation index images, like the commonly used NDVI. These images are then stitched together into a georeferenced aerial map of your field that provides much more than a pretty picture. Rather, they serve as an X-ray for your field, identifying problems you didn’t even know existed.

The uses for the color images are very straightforward, but still quite valuable. Agribotix customers have used color images from everything to identifying the presence of perennial weeds in a fallow field to catching clogged pivot spray heads to diagnosing drainage issues through catching wet spots.

This color image identified a clogged pivot sprayer head, enabling the farmer to clear the blockage before he would have noticed on his own and averting harm to that region of the field.

This color image identified a clogged pivot sprayer head, enabling the farmer to clear the blockage before he would have noticed on his own and averting harm to that region of the field.

While the color images are very useful for the gross diagnosis of farm problems, the infrared (NIR) images are much more illuminating. By subtracting the visible signal from the NIR signal, as discussed in depth in a previous blog post, a Difference Vegetation Index (DVI) image is generated. These are the unnormalized equivalents to the Normalized Difference Vegetation Index (NDVI) images that are commonly taken from the Landsats, but Agribotix research has determined that DVI processing is far superior for drone-collected imagery.

The NDVI image (left) and DVI image (right) are shown of a canopied barley field. Notice the NDVI image shows the healthiest area of the field to be in the upper half of the pivot close to the road and the DVI image shows just the opposite. Elsewhere in the fields these images give similarly conflicting results.

The NDVI image (left) and DVI image (right) are shown of a canopied barley field. Notice the NDVI image shows the healthiest area of the field to be in the upper half of the pivot close to the road and the DVI image shows just the opposite. Elsewhere in the fields these images give similarly conflicting results.

The conflict between these two image-processing techniques was initially concerning, but a long walk through the field confirmed the accuracy of the DVI image. The barley was totally headed out in the green areas of the DVI image displaying robust seed development. Conversely, the yellow, orange, and red areas of the field showed much weaker development and heads were not visible in most of those areas. The upper region of the field that the NDVI image identified as the healthiest was in fact the weakest in the field.

DVI image of canopied barley field clearly identifies stronger and weaker field zones. Notice the large, whitish barley heads in the image of the green area on the ground and the lack of development in the yellow and orange hole in the southern part of the field. In the weak northern area, the red finger that extends into the field corresponds to the greener areas in the image, where the barley has not begun to produce grain yet.

DVI image of canopied barley field clearly identifies stronger and weaker field zones. Notice the large, whitish barley heads in the image of the green area on the ground and the lack of development in the yellow and orange hole in the southern part of the field. In the weak northern area, the red finger that extends into the field corresponds to the greener areas in the image, where the barley has not begun to produce grain yet.

This field X-ray will be used to address the drainage issues causing these problems, guide top dress for the next growing season, and allow the grower to estimate his yields before harvest by identifying areas that will fail to produce.

However, Agribotix has extended the use of these DVI vegetation maps far beyond barley. Corn is king in agriculture and the ability to precisely top dress a field several times mid-season has the potential to both increase yields and reduce fertilizer input. While cornfields appear very homogeneous from the road, we have found dramatic variation in almost every cornfield we’ve flown or processed. The image below is very typical.

DVI image of this ca. 100-acre cornfield identifies dramatic heterogeneity that is not visible from the perimeter and was unknown to the grower.

DVI image of this ca. 100-acre cornfield identifies dramatic heterogeneity that is not visible from the perimeter and was unknown to the grower.

Magnifying the western portion of the field, it is clear that there are three general regions as identified by the DVI map. In the green band across the middle, the corn is nearly chest high (Tom is 6’4”), in the red and black band in the northwest, the corn is more like ankle high, and in the yellow band to the east the corn is about waist high.

Magnifying the western portion of the field, it is clear that there are three general regions as identified by the DVI map. In the green band across the middle, the corn is nearly chest high (Tom is 6’4”), in the red and black band in the northwest, the corn is more like ankle high, and in the yellow band to the east the corn is about waist high.

Because nitrogen was evenly applied to the field at the start of the season, the size of the corn is a direct proxy for the amount of nitrogen that has been pulled out of the soil. It is necessary to ensure that the corn has enough nitrogen to produce large kernels even after developing dense foliage, so they images can be converted into prescription maps that your variable rate applicator can read. Our customers have determined that binning the field into three zones, a 40 pound zone, 50 pound zone, and a 60 pound zone, will allow them to both reduce their overall fertilizer application and increase yields by providing the most fertilizer only where it has been depleted from the soil. Without these images, the entire field would receive 50 pounds and some parts would be under fertilized while others would be over fertilized.

A prescription map of the same field that can be imported into SST, SoilMaps, and most other precision ag software solutions. The field was binned into three zones for 60, 50, and 40 pounds/acre.

A prescription map of the same field that can be imported into SST, SoilMaps, and most other precision ag software solutions. The field was binned into three zones for 60, 50, and 40 pounds/acre.

Furthermore, while Agribotix imagery allow for accurate mid-season top-dress, it also allows the identification of weak zones that can be addressed at the end of the season prior to planting. In the barley field, the weak zones were related to drainage issues that the grower will seek to mitigate. In the cornfield, it is a combination of soil type, drainage, and likely a plugged nozzle.

Just as a physician had to rely on indirect evidence to diagnose a broken arm before the advent of the X-ray, prior to the advent of UAV-collected imagery, you as a farmer or agronomist had to rely on field walks, soil samples, and other sources of information. For the first time, you can get an accurate X-ray of your field that provides overall information about the health and function of the field, accurate prescription maps for precision top dress, and identification of soil problems that can be addressed before the next planting. Check out our lease program if you’re interested in bringing an Agribotix Hornet UAV onto your farm or Connect with Agribotix for more information.