Not a week goes by without a new article on use of drones for agriculture. A lot has been written. It is often said that it’s not how the data is collected but the software and what it can provide that is important. Yes, software is key. And yes, also the inputs, so the quality of the acquired data that is used in the software is of even higher importance. Why?
Garbage in, garbage out.
In other words, more sophisticated algorithms won’t deliver the right results if the information entered is not accurate. It’s not only the camera, or the data processing, or the UAV that are important, it’s the entire acquisition chain that has to be mastered at a high level. If only one particular link in the chain is weak, the entire chain will perform poorly.
So, when thinking about making decisions based on UAVs, or any other remote sensing platform, looking at the quality of the entire platform and its sensors will pay off.
This is the reason why investing in every detail is crucial.
First, let’s focus on the importance of the UAVs. To enable excellent conditions for aerial shot of agricultural fields, fixed wings needs to be incredibly stable. The two Delair’s UAVs UX5 and DT18 have been designed to enable picture acquisition with very few variations in the angle of acquisition, without blur, even in windy field conditions. This is also due to best in class sensors, with top level integration and communication of all the sensors and software that work smoothly together during flight. As a result, the quantity and the quality of images that can be acquired is, by far, the best in the market, fitting top level requirements of the ag industry: you get more than pictures – like with most of the UAVs – you get measurements. And measurements are required for most of the added-value applications.
Then, what about the camera? The Delair’s UAVs come with the MicaSense® Red Edge multi-spectral camera. It’s currently the best sensor for ag in the market, with industrial quality, very little distortion and noise – and the only one to acquire 5 distinct bands with this level of quality. In fact, its’s proven to no spectral contamination for the most accurate vegetation analysis. Its 5 narrow bands cover a wide range of crop parameters. First, we get accurate and valuable bird view pictures thanks to Red, Blue and Green. Red and blue bands are used to monitor the usage of energy of plants for photosynthesis. The Red edge band is sensitive to chlorophyll concentration which helps us to have a closer look at the stress on plant. Last, but not least, the Near Infra-Red, which is the most known band, is sensitive to biomass quantity and leaf cell structure.
On top of the camera, the use of a calibration panel (a white target with known behavior in the different bands) allows running a calibration process over each captured image, removing effects of light changing conditions and thus enabling accurate comparison over days and locations.
The Red Edge sensor is plugged into the UX5 and the DT18 aircraft triggering top level data acquisition of fields over more than 1000 hectares in a single flight, at the plant level.
So, taking care at every aspect of the drone’s data capture is essential for accurate measurements – and thus enable agriculture companies with plans to value drones on large scale and professional deployment. But then, once the risk of “Garbage in” has been removed, how to extract valuable insights and translate the data into action?
Some hints on the valorization of UAVs in the ag industry coming soon, stay tuned!
To be up to date, subscribe to our newsletter!
Lénaïc Grignard, agriculture product manager at Delair