Plant Imaging Diagnosis Using Leaf Counts for Stem Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for calculating the number of stems in plants, such as paddy rice, are hindered by the need for specific correlations between vegetation coverage rate and stems, which vary with cultivation conditions, and struggle to accurately determine stems when the coverage rate approaches 1.
Innovation Solution
Utilizing a correlation between the number of leaves and stems, specifically through image analysis with machine learning and neural networks to detect leaf tips, allowing for the calculation of stem numbers independently of planting density and growth stage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the vegetation coverage rate method is used to calculate the number of stems, then the calculation can be performed using image data, but the accuracy deteriorates when the vegetation coverage rate approaches 1 and varies according to different cultivation conditions
Solution Approach 1:
The patent changes the parameter used for calculation from vegetation coverage rate to the ratio of green pixels to total pixels in the plant region. This parameter change allows accurate stem number calculation even when vegetation coverage is high, as it directly measures the proportion of active plant tissue rather than overall coverage area.
Solution Approach 2:
The patent segments the image processing into distinct steps: first identifying the plant region, then calculating the ratio of green pixels to total pixels within that region, and finally deriving the stem number from this ratio. This segmentation allows each step to be optimized independently, improving overall accuracy.
2Productivity
If the vegetation coverage rate method is used, then stem number can be acquired based on correlation, but the method requires acquiring separate correlations for each cultivation condition increasing complexity
Solution Approach 1:
The patent creates a universal calculation method that works across different cultivation conditions without requiring condition-specific correlations. The method uses the ratio of green pixels to total pixels in the plant region, which is a universal metric that maintains consistent relationship with stem number regardless of planting density or growth stage.
Solution Approach 2:
Instead of creating separate correlation models for each cultivation condition, the patent copies and applies the same calculation methodology (green pixel ratio method) across all conditions, simplifying the system while maintaining accuracy.
Data Source
AI summary
An imaging diagnosis apparatus includes a first acquiring unit configured to acquire information on the number of leaves of a plant using an image of the plant, and a second acquiring unit configured to acquire the information on the number of stems of the plant using the information on the number of leaves.