Cotton Fiber Image Classification for Short Fiber Measurement

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for measuring and classifying cotton fibers are inaccurate, particularly in determining the length of short fibers, leading to inefficiencies in processing and increased contamination, and lack the ability to provide a quantitative and qualitative analysis of fiber characteristics.

Innovation Solution

A method involving image processing techniques to analyze fiber layers, applying filters to acquired images, segmenting fibers into groups, and calculating lengths and characteristics, allowing for precise classification and characterization of fibers.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional beard sample preparation method is used, then the measurement process is simple, but the measurement precision of fiber length is poor and only estimated quantitative analysis is obtained

Engineering Contradiction:
Improvefiber length measurement precisionVSAvoidimage processing system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces the mechanical manual preparation and estimation method with an automated image processing system using digital cameras and computer algorithms to automatically measure fiber length and characteristics, achieving precise measurement while eliminating manual intervention

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent creates a digital copy (image) of the physical fiber sample, then analyzes this copy through image processing to determine fiber characteristics without physically manipulating the original sample, enabling precise non-contact measurement

Inventive Principle:
Principle #26Copying

2Productivity

If short fiber proportion increases due to mechanical picking and hard ginning, then productivity is improved, but measurement accuracy of short fibers deteriorates and contamination increases

Engineering Contradiction:
Improvecotton processing productivityVSAvoidshort fiber measurement accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent changes the measurement parameters by using image processing algorithms that can detect and measure fibers of all lengths including very short fibers, with the ability to set customizable length thresholds (e.g., 10mm, 11mm, 13mm, 13.5mm) to accurately determine short fiber proportion regardless of how low the absolute value is

Inventive Principle:
Principle #35Parameter changes

3Productivity

If a large proportion of short fibers is present, then productivity may be maintained, but fly contamination and strain on personnel and machines increases

Engineering Contradiction:
Improvefiber processing outputVSAvoidfly contamination
Core Design Contradiction:
ProductivityVSObject-generated harmful factors

Solution Approach 1:

The patent provides accurate feedback on the actual short fiber proportion and distribution through automated image analysis, enabling operators to make informed decisions about processing parameters and contamination control measures based on precise quantitative data rather than estimates

Inventive Principle:
Principle #23Feedback

4Ease of operation

If standard short-fiber limit is set at 13 or 13.5 mm, then classification is simplified, but measurement accuracy deteriorates because short fibers cannot be measured easily

Engineering Contradiction:
Improveclassification simplicityVSAvoidshort fiber length measurement precision
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent performs preliminary action by capturing a complete digital image of all fibers in the sample before measurement, allowing the system to analyze and measure fibers of any length including those at the standard threshold (13mm or 13.5mm) with high precision through automated image processing rather than manual measurement

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4678301A1Method of cotton fiber classification
Publication Date: 2026.01.14 RIETER CZ AS
  • EP4678301A1 patent drawingFigure 1~2
  • EP4678301A1 patent drawingFigure 3~7
  • EP4678301A1 patent drawingFigure 8

AI summary

The present invention concerns a method (500) of retracing of fiber in a fiber layer (900). Said fiber layer (900) is spread on a surface such as to acquire (510) an image (511) of said fiber layer (900) comprising a plurality of fibers (910) including at least one fiber (911). At least one image-filter (531) is applied to said acquired image (511) and a at least one list (120) is created (570) per said at least one fiber (911) such as to classify said fiber.