Crop Canopy Imaging Fusion for Field-Scale Growth Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The scarcity of comprehensive and detailed earth observations hinders the understanding of terrestrial ecosystem change, as conventional remote sensing methods provide limited spatio-temporal data, and ground camera-collected vegetation data is considered point-level, failing to capture field-level variations.
Innovation Solution
Integrating Internet-of-Things (IoT) devices, multi-satellite images, and process-based models to collect and upscale environment and vegetation data from ground and space, generating high-resolution crop growth condition maps in real-time for accurate crop productivity estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If ground cameras are used to collect vegetation data, then continuous observations of canopy structure and leaf color are provided, but the data can only be considered as point level and cannot capture field-level variations
Solution Approach 1:
The patent transitions from point-level ground camera observations to field-level coverage by integrating satellite remote sensing data. This adds the spatial dimension of satellite imagery to the temporal and spectral dimensions provided by ground cameras, enabling both continuous monitoring and field-level representativeness simultaneously.
Solution Approach 2:
The patent merges ground-based camera observations with satellite remote sensing data to create a comprehensive monitoring system. By combining the continuous temporal data from ground cameras with the spatial coverage of satellite imagery, the system achieves both measurement precision and spatial representativeness.
2Adaptability or versatility
If remote sensing is used to capture spatial heterogeneity, then field-level vegetation data is obtained, but the data requires complex models to convert signals to ecosystem variables and lacks comprehensive ground truth data
Solution Approach 1:
The patent uses ground camera observations as an intermediary to validate and calibrate satellite remote sensing models. By having ground truth data from cameras, the system can develop and refine algorithms to accurately convert satellite signals into ecosystem variables, reducing the complexity of modeling requirements.
Solution Approach 2:
The system incorporates feedback loops where ground camera data is used to continuously improve and refine satellite-based models. This iterative process allows the models to adapt to local conditions and improve their accuracy in converting remote sensing signals to ecosystem variables.
3Reliability
If comprehensive ground truth data is collected at multiple experimental sites during limited time periods, then accurate remote sensing models are developed, but the high cost of data collection limits the scope and duration of observations
Solution Approach 1:
The patent employs autonomous ground cameras that continuously self-monitor vegetation conditions without requiring extensive manual data collection. These cameras automatically capture images and transmit data, enabling long-term observations at multiple sites while reducing the resource-intensive processes of manual data gathering and site visits.
Solution Approach 2:
The ground camera system serves multiple functions: it provides ground truth data for model validation, monitors canopy structure over time, and captures spatial variations across different sites. This multi-functionality allows one system to address multiple research needs simultaneously, expanding the scope of data collection without proportionally increasing costs.
Data Source
AI summary
One or more cameras or other image collecting devices are used at or near the surface of a field including crops to acquire images of the crop canopy. This can be done by tilting the cameras relative to the zenith. Additional imagery, such as airborne or satellite of the same crop canopies is also acquired, with at least some of the imagery having similar geographical and time-stamped information. A machined-learning model will take the near surface imagery and be trained to apply the information to the remote imagery of the air or satellite to be applied to a greater area, which provides plant growth information to the wide area of the remote imagery.