Plantation Image Recognition With Adaptive Two-Stage Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Agricultural machines face challenges in real-time image recognition of crops, weeds, insects, and pathogens due to varying illumination and phenotypical changes, which affects the accuracy of treatment decisions in plantation fields.
Innovation Solution
A method that involves running a first image recognition analysis with initial parametrization, identifying unsatisfying results, and then using ambient data to run a second, more complex analysis on an external device to improve the parametrization for the machine learning algorithm, thereby enhancing the accuracy of image recognition and controlling treatment devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a first image recognition analysis of lower complexity is used for real-time treatment decisions, then processing speed is improved, but detection accuracy deteriorates
Solution Approach 1:
The image recognition process is divided into two separate analyses: a first simpler analysis for real-time processing decisions and a second more complex analysis for accurate identification. This segmentation allows each analysis to be optimized for its specific purpose, resolving the contradiction between speed and accuracy.
Solution Approach 2:
The first image recognition analysis acts as an intermediary that prepares initial treatment decisions based on speed-optimized processing, while the second analysis serves as a mediator to refine and verify these decisions with higher accuracy, combining the benefits of both fast and accurate processing.
2Measurement precision
If a second image recognition analysis of higher complexity is used to improve detection accuracy, then detection accuracy is improved, but processing time increases
Solution Approach 1:
The first image recognition analysis is performed as a preliminary action to quickly identify items requiring treatment and make initial treatment decisions. This preliminary processing reduces the overall time burden by handling routine cases rapidly, allowing the more time-consuming second analysis to focus only on complex or uncertain cases.
Solution Approach 2:
Instead of applying the complex second analysis to all images, the system uses partial action by applying it selectively only when needed to improve or verify treatment decisions. This reduces the total processing time while still achieving high accuracy where necessary.
3Productivity
If image recognition algorithms are used to treat plantation in real-time, then productivity is improved, but reliability of treatment decisions deteriorates due to heterogeneous conditions
Solution Approach 1:
The system uses feedback by comparing results from both image recognition analyses and using the second analysis to verify and improve treatment decisions. This feedback mechanism increases reliability by catching errors that the first analysis might miss, while maintaining productivity through the efficient two-stage process.
Solution Approach 2:
The system changes parameters by adjusting the complexity level of image recognition analysis based on the specific conditions and requirements of each treatment decision. This allows the system to optimize between speed and reliability dynamically, improving overall reliability without sacrificing productivity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Method for plantation treatment of a plantation field, the method comprising taking an image of a plantation of a plantation field; recognizing items on the taken image by running a first image recognition analysis of a first complexity on the taken image based on a stored parametrization of a machine learning algorithm; identifying an unsatisfying image analysis result; determining ambient data corresponding to the taken image; recognizing items on the taken image by running a second image recognition analysis of a second complexity on the image based on the ambient data on an external device, wherein the second complexity is higher than the first complexity; determining an improved parametrization based on the second image recognition analysis for the machine learning algorithm for improving the first image recognition analysis; and controlling a treatment arrangement of a treatment device based on the first image recognition analysis.