The Process
Discover how we use drones, hyper-spectral imaging, data analytics, and AI
Flight Tests
Ag Genius AI utilizes fixed-wing drones capable of operating beyond visual line of sight (BVLOS) at a height of 1,000 feet above ground level. We integrate high-resolution hyper-spectral aerial images captured by drones, covering a range of 20,000 to 40,000 acres daily, with lower resolution satellite data that captures multispectral information over much larger regions.
The primary objective is to assess the economic viability, scalability, automation, and data science requirements essential for consistently delivering imaging services on a commercial scale. Ag Genius AI conducts flights multiple times per week across varied soil types, capturing numerous photos of crops and fields throughout their growth stages.
With a promising forecast of achieving costs below $0.50 per acre per day compared to the current range of $2 to $6 per acre per day, Ag Genius AI is poised to revolutionize the accessibility and affordability of aerial imaging in the agricultural sector.
Ag Genius AI compares images within each growth stage to demonstrate the viability of forecasting:
Crop conditions
Yields
Commodity pricing
Vegetative indices
Other indicators
Cultivating Data
Vegetative Indices We Will Compute
Normalized Difference Vegetation Index (NDVI)
Indicates vegetation greenness and canopy density
Enhanced Vegetation Index (EVI)
Also for greenness but less sensitive to background and atmosphere
Advanced Soil Adjusted Vegetation Index (SAVI)
Modified NDVI for reducing soil influences
Modified Soil Resistance Vegetation Index (MSRAVI)
Alternative soil adjusted index
Anthocyanin Reflectance Index (ARI)
Pigmentation indicator for plant stress
Water Band Index (WBI)
Canopy water content measurements
Triangular Chlorophyll Index (TCI)
Chlorophyll estimates with leaf angle normalization
Pigment Specific Simple Ratio (PSSR)
Carotenoid estimator for excess light stresses
Gitelson Model 1 (GM1)
Invertase activity measure related to sucrose accumulation
Normalized Difference Red Edge Index (NDRE)
Chlorophyll absorption indicator
Photochemical Reflectance Index (PRI)
Light use efficiency and CO2 uptake indicator
Moisture Stress Index (MSI)
Leaf water content index
Modified Chlorophyll Absorption Ratio Index (MCARI)
Chlorophyll concentration with less background influences
Red Edge Chlorophyll Index (RCC)
Early signs of crop diseases
Global Environmental Monitoring Index (GEMI)
Vegetation health assessment using curve modeling
Green Chlorophyll Index (CIgreen)
Estimates leaf chlorophyll content
Structure Intensive Pigment Index (SIPI)
Carotenoid to chlorophyll ratio
Normalized Difference Infrared Index (NDII)
Leaf water evaluator
Plant Senescence Reflectance Index (PSND)
Pigment change assessor as plants age
Blue Wide Dynamic Range Vegetation Index (BWDRVI)
General stress and pest detector
The Fusion Process
Studies have shown these fusion techniques can improve classification accuracy and provide a more consistent time-series when combining data sets from multiple platforms and sensors.
-
The high-resolution hyperspectral data from the drone can be used to sharpen or up sample the lower resolution multispectral satellite imagery using resolution merge techniques like pansharpening. This enhances the spatial resolution of the satellite data to better match the drone data.
-
The differing spectral responses between the sensors can be normalized using spectral convolution or empirical line calibration methods. This puts the data into the same spectral space.
-
Advanced techniques like coupled non-negative matrix factorization, Bayesian data fusion, or stacked autoencoders can be used to integrate the hyperspectral and multispectral data into a unified data cube. This leverages the spectral information from the hyperspectral data and the synoptic coverage from the satellite multispectral data.
-
The output maps, classifications or data products from the separate hyperspectral and multispectral analysis can be fused using techniques like weighted averaging, Evidential reasoning, or random forest-based integration. This combines the outputs from the different data streams.
Techniques
Collect multi-spectral drone images over selected benchmarked areas that also have existing multispectral satellite coverage.
Benchmark zones of high-resolution drone images and lower resolution satellite images will provide baseline data sets for a controlled comparison.
Benchmark areas will be processed using our proprietary data fusion techniques to integrate the multispectral satellite data with the hyperspectral drone images.
The Benefits
The rapid advancements in drone technology have ushered in a new era of precision agriculture, empowering farmers to harness the power of aerial data and imagery to transform their farming practices. Drones, equipped with a suite of advanced sensors, have become indispensable tools in the modern agricultural landscape, offering a multitude of benefits that can significantly improve crop yields.
-
Drones equipped with multispectral and RGB cameras can capture detailed information about the overall health and vigor of crops, allowing farmers to identify potential issues early on. By analyzing vegetation indices, such as the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Red Edge Index (NDRE), farmers can detect nutrient deficiencies, water stress, and disease outbreaks, enabling them to take targeted, preventive action.
-
Drone-based thermal imaging and multispectral data can be used to map soil moisture levels and identify areas of the field that require more or less water. This information can be integrated with precision irrigation systems, allowing farmers to optimize water usage and ensure that each part of the field receives the appropriate amount of water, ultimately improving water-use efficiency and crop yields.
-
Drones can be equipped with sensors that measure the spatial variability of soil properties, such as nutrient levels and pH, across a field. This data can then be used to create variable-rate application maps, enabling farmers to precisely apply fertilizers, pesticides, and other inputs based on the specific needs of different areas of the field, thereby reducing waste and maximizing the efficacy of these inputs.
-
Drones equipped with high-resolution cameras can capture detailed imagery of crops throughout the growing season, allowing farmers to monitor crop development and estimate yields more accurately. This information can be used to plan harvesting operations, optimize storage and logistics, and even forecast future production levels, enabling better decision-making and resource allocation.
-
Drones equipped with high-resolution cameras and multispectral sensors can quickly identify early signs of pests, diseases, and environmental stresses, such as nutrient deficiencies or water logging, that may not be readily visible to the naked eye. This early detection allows farmers to implement targeted interventions, such as spot-spraying of pesticides or precision application of fertilizers, before the problems escalate and cause substantial crop damage.
-
Drones can be used to conduct efficient and comprehensive field scouting, covering large areas in a shorter amount of time compared to traditional manual methods. This enables farmers to identify and address issues more quickly, reducing the risk of further crop losses. Additionally, drones can access areas of the field that may be difficult or dangerous for ground-based scouting, providing a more complete picture of the crop's condition.
-
By repeatedly flying drones over a field and capturing high-resolution imagery, farmers can closely monitor the growth and development of their crops throughout the growing season. This information can be used to identify growth anomalies, plan for harvest, and make informed decisions regarding the timing and application of inputs, such as fertilizers and irrigation.
-
Drones can be equipped with precision spraying systems that allow for the targeted application of pesticides, herbicides, and other crop protection products. This enables farmers to treat specific problem areas within a field, reducing the overall amount of chemicals used and minimizing the risk of environmental contamination.
-
As mentioned earlier, drone-based thermal imaging and multispectral sensors can be used to map soil moisture levels and identify areas within a field that require more or less water. This information can be integrated with precision irrigation systems, enabling farmers to optimize water usage and ensure that each part of the field receives the appropriate amount of water, thereby reducing water waste and improving crop yields.
-
Drones equipped with sensors that can measure the spatial variability of soil properties, such as nutrient levels and pH, can help farmers create variable-rate application maps. This allows for the precise application of fertilizers, based on the specific needs of different areas within the field, reducing waste and maximizing the efficacy of these inputs.
-
Drones can quickly identify areas of the field affected by pests, diseases, or weed infestations, enabling farmers to target their pesticide and herbicide applications only to the affected areas. This precision approach can significantly reduce the overall amount of chemicals used, leading to cost savings and mitigating the environmental impact of these inputs.
-
Drones can be used to conduct comprehensive field surveys and monitor crop growth, reducing the need for extensive manual scouting by farm workers. This frees up labor resources that can be redirected to other critical tasks, such as targeted weed removal, pest control, or harvesting operations, thereby improving the overall efficiency and productivity of the farm.
-
Drones equipped with high-resolution cameras and multispectral sensors can be used to detect the early signs of pest infestations, such as discoloration or defoliation of leaves, before they become widespread. This early warning system allows farmers to implement targeted pest control measures, reducing the risk of significant crop damage and the need for broad-spectrum pesticide applications.
-
Similarly, drone-based imagery and data can be used to identify the early stages of plant diseases, such as fungal infections or bacterial blights, by detecting changes in leaf color, texture, or growth patterns. Early disease detection enables farmers to take prompt action, such as applying fungicides or implementing cultural practices, to prevent the spread of the disease and minimize its impact on crop yields.
-
Drones equipped with multispectral cameras can capture data on the reflectance and absorption characteristics of crop leaves, which can be used to identify nutrient deficiencies or imbalances. This information allows farmers to supplement the affected areas with targeted fertilizer applications, ensuring that their crops receive the necessary nutrients for optimal growth and productivity.
-
Drones can also be used to detect and monitor various environmental stresses that can impact crop performance, such as water logging, drought, or temperature extremes. By identifying these stress factors early on, farmers can implement appropriate mitigation strategies, such as adjusting irrigation schedules, applying shade nets, or implementing protective measures, to minimize the negative effects on their crops.
-
Drones equipped with high-resolution cameras and advanced sensors can provide farmers with detailed, real-time data on crop growth, development, and potential yield (Berni et al., 2009; Yao et al., 2017). This information can be used to make informed decisions regarding the timing and application of inputs, such as fertilizers, pesticides, and irrigation, to optimize crop productivity. By addressing the specific needs of different areas within a field, farmers can maximize yield potential and improve overall crop performance.
-
The precision-based approach enabled by drone technology can lead to significant reductions in the amount of inputs, such as water, fertilizers, and pesticides, required to achieve optimal crop yields. By targeting the application of these inputs to specific areas of the field, based on the data collected by drones, farmers can minimize waste and maximize the efficacy of these resources, resulting in cost savings and improved resource-use efficiency.
-
Drone-based monitoring and management can also contribute to improved crop quality, as farmers are able to identify and address issues, such as nutrient deficiencies or pest infestations, at an early stage. This can lead to higher-quality produce with better market value, further enhancing the profitability of the farming operation.
-
By enabling early detection of problems and facilitating more targeted interventions, drone technology can help farmers minimize crop losses due to pests, diseases, or environmental stresses. This reduction in wastage translates directly into increased yields and higher overall productivity for the farm