Soybean Quality Assessment Using Machine Vision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for assessing soybean quality are subjective, prone to variance, and lack automation, particularly in rural and remote environments, leading to inconsistencies in grading and evaluation based on visual factors like texture, shape, and color.
Innovation Solution
An automated system employing machine vision and Haralick texture analysis, combined with shape and color analysis, uses a custom-made imaging plate and LED lighting for high-resolution imaging and processing, capable of distinguishing between smooth, cracked, and wrinkled soybeans, and deployed in rural settings for accurate and repeatable evaluation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual inspection and grading based on Visual Reference Images is used, then inspectors can evaluate soybean quality using established USDA standards, but the process suffers from subjective variation, inspector experience differences, and inconsistent grading across different locations
Solution Approach 1:
The patent replaces the manual mechanical inspection process with an automated machine vision system that captures images of soybeans and uses computer algorithms to objectively assess quality factors including texture, shape, and color, eliminating inspector subjectivity and experience-based variation
Solution Approach 2:
The system creates digital copies (images) of soybean samples and uses these copies for analysis instead of requiring physical inspection, allowing multiple analyses of the same sample without degradation and enabling automated processing
2Productivity
If automated image processing methods are implemented, then inspection speed and consistency improve, but the systems lack widely-accepted criteria for defining normal versus damaged soybeans, leading to potential misclassification
Solution Approach 1:
The patent transforms the classification problem from binary (normal/damaged) to multi-dimensional by analyzing multiple parameters simultaneously (texture features, shape metrics, color values) and comparing them against established USDA grading criteria, enabling more nuanced and accurate classification
Solution Approach 2:
The system incorporates feedback mechanisms where classification results are continuously refined by comparing against known standards and adjusting algorithms to improve accuracy over time, ensuring reliability while maintaining high processing speed
3Ease of manufacture
If simple binary classification methods are used (damaged vs. normal), then the system is easy to implement, but it incorrectly classifies small and irregular but otherwise undamaged soybeans as damaged
Solution Approach 1:
The patent segments the quality assessment into multiple independent analysis components (texture analysis, shape analysis, color analysis) rather than using a single binary classification, allowing each aspect to be evaluated separately and combined for a comprehensive quality determination that reduces false positives
Data Source
AI summary
A method and apparatus for evaluating soybean quality are disclosed. Automated or semi-automated image processing can be implemented with commercially available machine vision hardware. Post-harvest soybeans are separated, imaged, and analyzed to assess grain quality factors used in official grading. The system may be used prior to sale, purchase, further harvesting, or planting subsequent crops. The system reduces inspection errors, reduces variance in results, increases efficiency, enhances repeatability, and improves the standardization of soybean quality evaluation. Shape analysis, color analysis, and Haralick texture analysis are employed to determine when certain types of damage are present, to distinguish between types of damage, and to classify grains as smooth, cracked, or wrinkled. The system is sufficiently robust to allow successful deployment in rural, remote, and non-standard environments.