Seed Coating Image Analysis for Consistent Abrasion Assessment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for assessing the abrasion resistance of seed coatings are subjective and lack consistency due to uncontrolled random factors, leading to high variation in determining the quality of film coatings on seeds.
Innovation Solution
A method utilizing a trained neural network to analyze images of seeds, determining the coverage and uniformity of coatings by processing images of seeds exposed to mechanical stress, employing a convolutional neural network with layers such as input, 2D convolution, batch normalization, ReLU, max pooling, and regression to accurately assess abrasion resistance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If visual inspection methods are used to assess abrasion resistance, then the assessment process is simple and quick, but the results are subjective and lack consistency due to uncontrolled random factors
Solution Approach 1:
The patent replaces the manual visual inspection process with an automated image processing system that captures images of seeds and uses computational algorithms to objectively measure coating coverage and uniformity. This substitution of mechanical/manual assessment with automated imaging technology eliminates human subjectivity while maintaining operational simplicity.
Solution Approach 2:
The patent introduces image processing algorithms and computational models as intermediaries between the physical seed coating and the final assessment result. These intermediaries objectively quantify coating properties by analyzing image data, thereby eliminating the inconsistency inherent in direct visual inspection while keeping the overall process straightforward.
2Measurement precision
If automated image processing with neural networks is used to assess abrasion resistance, then the measurement precision and consistency are improved, but the device complexity increases
Solution Approach 1:
The patent employs trained neural networks that automatically learn and extract relevant features from seed images without requiring manual feature engineering or complex preprocessing. The system self-calibrates through training on labeled data, reducing the need for complex system configuration and maintenance while achieving high measurement precision.
Solution Approach 2:
The patent performs preliminary training of neural networks using labeled image data before deployment. This preliminary action prepares the system in advance, allowing it to automatically and accurately assess abrasion resistance without requiring complex real-time adjustments or interventions during actual measurement, thereby reducing operational complexity.
3Adaptability or versatility
If multiple training images with different coating colours are used to train the neural network, then the adaptability of the system is improved, but the training time and data requirements increase
Solution Approach 1:
The patent trains a single universal neural network model on diverse training data encompassing multiple coating colours and types. This universal model learns to generalize across different coating characteristics, enabling it to accurately assess various seed coatings without requiring separate specialized models for each colour, thereby reducing overall training time and computational resources.
Data Source
Figure 1a~1b
Figure 2
Figure 3
AI summary
A method for processing images of seeds, the method comprising: inputting a seed image comprising at least a portion of a seed to a trained neural network to generate a value associated with the coverage of a coating on the seed; the trained neural network having been trained to generate a value associated with the coverage of coating on the seed using a plurality of training images each comprising at least a portion of a training seed, and wherein each training image is labelled with a value associated with the coverage of a coating on the training seed.