Kernel Component Analysis via Spectroscopy Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for analyzing kernel components, such as rice quality evaluation, lack accuracy and efficiency, particularly when analyzing on a kernel-by-kernel basis, and require destructive or time-consuming processes.
Innovation Solution
A kernel component analysis device and method using spectroscopy that includes a light emitter, spectrum detector, and computing unit to quantify specific components in kernels on a kernel-by-kernel basis, employing a calibration curve to calculate component content from detected spectrum values, with improved measurement accuracy by focusing on a suitable effective portion of the kernel image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If near-infrared spectroscopy is used for component analysis, then nondestructive and real-time examination is achieved, but measurement accuracy on a kernel-by-kernel basis is insufficient
Solution Approach 1:
The patent divides the kernel image into multiple regions of interest (ROIs), including a central portion and peripheral portions, to selectively analyze specific areas. This segmentation allows the system to focus on regions with higher light transmission for improved measurement accuracy while maintaining nondestructive and real-time examination capabilities.
Solution Approach 2:
The patent applies different analysis methods to different regions of the kernel image. The central portion is analyzed using one method while peripheral portions are analyzed using another method, with each region's characteristics considered. This local quality approach improves measurement accuracy by tailoring the analysis to the specific properties of each kernel region.
2Measurement precision
If visual inspections or crushing kernels is used for quality evaluation, then component analysis is achieved, but time consumption and destructive processing occur
Solution Approach 1:
The patent replaces mechanical crushing and manual visual inspection with optical spectroscopy. By using near-infrared light to penetrate and analyze kernel components nondestructively, the system eliminates time-consuming mechanical processing while maintaining or improving measurement accuracy through automated spectral analysis.
Solution Approach 2:
The kernel itself serves as the sample for analysis, requiring no external preparation or destruction. The near-infrared spectroscopy system directly analyzes the kernel's optical properties to determine component content, making the kernel itself the source of information without needing to be crushed or prepared in any way.
3Area of stationary object
If light is transmitted through the entire kernel image, then complete coverage is achieved, but measurement accuracy decreases due to varying light attenuation
Solution Approach 1:
The patent segments the kernel image into a central portion and peripheral portions, analyzing each region separately. The central portion, which has higher light transmission and better spectral characteristics, is prioritized for quantitative analysis, while peripheral portions are analyzed with adjusted parameters to compensate for light attenuation.
Solution Approach 2:
The patent applies different weightings and analysis methods to different regions of the kernel image based on their local quality characteristics. Regions with better light transmission (central area) receive higher weighting for quantitative analysis, while peripheral regions are analyzed with adjusted parameters to maintain overall measurement accuracy.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables nondestructive, real-time, and accurate measurement of kernel components, enhancing the precision of protein content analysis and quality evaluation on a kernel-by-kernel basis, with the potential for automated screening and classification.
Implementation Method 1
a light emitter configured to irradiate a kernel to be analyzed with light; a spectrum detector configured to detect a spectrum of light transmitted through and/or reflected from the kernel irradiated with the light
Implementation Method 2
detecting a spectrum of light transmitted through and/or reflected from the kernel irradiated with the light
Data Source
AI summary
A kernel component analysis device quantitatively analyzing a specific component contained in each of kernels on a kernel-by-kernel basis by spectroscopy includes: a light emitter configured to irradiate a kernel to be analyzed with light; a spectrum detector configured to detect a spectrum of light transmitted through and/or reflected from the kernel irradiated with the light; and a computing unit configured to calculate, on a kernel-by-kernel basis, a content of the specific component in the kernel to be analyzed, based on a spectrum value detected from an effective portion, which is suitable for quantitative analysis, of an image of the kernel by using a calibration curve indicating a relationship between a spectrum value at a specific wavelength and a content of the specific component.