Aerial Image Analysis
DroneScope service for land area analysis based on aerial photography and computer vision
Aerial image analysis enables users to detect and analyze objects via footage taken by airborne platforms, such as fixed-wing aircraft, helicopters, unmanned aerial vehicles (UAVs or "drones"), balloons, etc. It is utilizing aerial views and computer vision technology to gather data from vast land areas and process it in a structured and automatic way. Adaptable for both small and large-scale applications, aerial image analysis can change the way private and public sectors gather information and manage landscapes.
Aerial image analysis has shown to be very useful in various industries. While each specific case might need a dedicated algorithm training process, aerial imagery analysis could be used in:
- agriculture for crop and yield management
- animal husbandry for cattle and field observation
- forestry for flora and fauna monitoring
- urbanization for city zone and power line planning
- transportation for road and rail maintenance
- other land area observation applications
New opportunities in land observation opened up due to advancements in drone technologies over the past few years. Aerial photography became flexible in application and more available to consumers. With more than 20 years of experience in the field of computer vision, Neurotechnology wanted to provide a technology that would enable users to apply aerial photography for automated land area monitoring and management. Therefore, the latest algorithms were used alongside the SentiSight.ai platform to develop new aerial image analysis solutions. One of the latest aerial image processing solutions Neurotechnology participated in developing is Dronescope.
DroneScope service and solution
Neurotechnology partnered with SmartField, a company specialized in planning and managing agriculture in South America, to create a new solution - DroneScope. The solution combines advanced computer vision and drone technologies to help the agriculture industry view, analyze and manage crops like never before.
Image acquisition and processing is not always an affordable technology. Agricultural drones are usually expensive to operate and require hiring specially trained operators, as well as using additional image processing tools to analyze the obtained media.
To provide more flexibility, the DroneScope solution was designed to use RGB images obtained with regular commercial drones. Drones below 900 gr with 12 Megapixel cameras and good object avoidance systems are perfectly suitable. Therefore, the DroneScope is suitable for both small and large-scale applications.
The aerial imagery recognition algorithm allows seeing crops sooner and closer than a satellite and provides 100 times better survey capacity than humans in the fields, and it also remains impartial during sampling. The solution enables agricultural users to monitor:
- Crop coverage – gather a detailed analysis of the coverage for each row of crops and the variability on the plot.
- Plant count – observe crop productivity by estimating the number of plants and how well distributed they are.
- Weeds coverage – determine and evaluate weeds spread on the plot.
- Plant diseases – evaluate the disease incidence.
- Maps, zoning, and prescriptions – use obtained imagery, weed maps, crop structure, and plant population to define the best spray strategy.
- Sentinel-1 satellites data processing – allows for analysis of NDVI, NDRE, MSAVI, RECI indexes, and NDMI humidity index. Also, allows monitoring of the daily history of ground temperature and humidity.
For farmers, consultants or growers, the DroneScope solution offers a reliable platform to process, store and share not only images but also the actual condition of their crops. It becomes an outstanding tool for immediate decision-making and a source of knowledge for the upcoming seasons.
Dronescope solution is provided as a service for farmers who would like their agricultural fields to get analyzed. Also, the DroneScope can be provided as an algorithm for integrators who are developing software systems for agriculture.