Have you ever heard about “cuscuta campestris” also known as “dodder weed”? We no, until a team of SPH Engineering joined GeoScan IL in their project for beetroot field inspection. Here is a photo how the dodder weed looks like:
Test field was planted with sugar beet and occupied around 12 Ha. It took less than an hour to capture all the images with a drone. After 2 hours we had a bird-view orthophoto map of the field at very high resolution. And in the next 1 hour, we had the following detection result built-in ATLAS.
As you can see from 10 to 20 percent of the field is covered with field dodder. There are several large areas, but also many small spots which are also very important to catch in a timely manner to prevent further spread. Any existing manual methods can not bring comparable results in terms of speed, accuracy, and visibility.
Weeds are dangerous for the following crops: sugar-beet, melons, asparagus, tomatoes, and others. And it is very important to control not only fields but also roadsides nearby and prevent any spread. The untimely maintenance of the field leads to the loss of crops and loss of revenue. Maintenance of the field includes getting rid of weeds, spraying from pests and diseases. And overuse of chemicals leads to increased costs.
Roman Kirsanov, CEO and Founder at GeoScan IL says: “The usage of ATLAS solution makes the process of POI identification and marking simpler than ever due to user-friendly interface, as well as helps to process large areas in short time with significantly less human time investment in the process. As a result, our workflow in the field survey became faster and more efficient than ever.”
Customer: GeoScan IL (Israel)
Industry: Agriculture, Drone inspection, Data processing
Process:
- DJI M600 with Agrowing's A7Rii-based multispectral sensor.
Map storage and processing in ATLAS:
- Manual annotation: approx. 10-15 minutes.
- Initial machine learning detector training in ATLAS: approx. 30 minutes
- Survey mission planned with UgC
Result:
- Shareable web-map with location and size of areas with weed.
- Report with georeferenced areas in GeoJSON and Shapefiles.
Using ATLAS, the customer receives a more accurate marking of fields for weeds and diseases.
The value can be measured by:
- savings on time of using the machinery (spraying from pests and diseases),
- savings on the volume of pesticides used,
- saving more crops from weed spread and damage.