Aero Hawk makes a very physically capable drone, but it’s also very smart. After all, anyone can develop a quadcopter with some cameras on it and a decent battery these days. What truly makes Aero Hawk’s drones special is their software.
Aero Hawk wanted to give members of the agricultural industry images that make data actually usable. This means that the software must be intuitive to use and also provide the correct non-diluted pixel placement.
The images taken of crops by aerial drones are initially separate. A drone’s software will then combine those photos to give a wider picture. This bigger, more comprehensive images allows for patterns and problems to be easily and more accurately identified than they would be in separate images of the same field or fields.
Most software on the market for agricultural drones uses a technique called blending or stitching. This process simply overlays the pictures on top of each other. This manipulation, however, creates new visual data in the area where the overlay is done. This can cause a loss of important information or inaccurate images of key areas. Something important might be missed.
Aero Hawk’s software is different. Using a concept referred to as the “Most Important Pixel”, developed by Aero Hawk’s Dr. Hermilo Hernandez-Noyola, Aero Hawk does not just stitch images together in the traditional way. In order to provide a more accurate picture of the plants, their software overlays the images, then pulls the visual data based on the “Most Important Pixel” (the center one) to give a more accurate image.
Let’s use an example. If you have between 120-1000 images of a 160-acre field. Half of those pictures are blended together with the neighboring ones. Those blending or stitch lines cause distortion in the images. With an 80% overlap of pictures, it can be a wide enough line that the whole image is distorted. This leaves substantial room for error if you are looking for signs of disease in a field. You are not very likely to be able to identify one plant with a problem since every plant is around 6 pixels in an 8 million pixel image. The distorted area could contain the signs of disease you’re looking for and you’ll never know.
Aero Hawk’s software avoids this issue by keeping the integrity of every image, using geo-referenced tie points and the “Most Important Pixel”. Because the software does not use blending to combine images, there is less likelihood of you losing important images of diseased plants or other problems in your fields.
Aero Hawk’s software is designed to be easy to use and truly helpful to the modern agricultural industry. It is continually being updated. Coming releases will include actionable data tools that help with decision making. As the industry and technology develop, the Aero Hawk team plans on improving their software to better help users identify issues in their fields sooner.