PCB Inspection Image Processing Using RGB Subtraction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image processing devices require significant upfront setting of RGB values for recognition targets and exclusion areas, which can lead to inaccuracies due to changes in lighting conditions or object orientation during image capture.

Innovation Solution

The image processing device generates a difference image based on the subtraction of gradation values from primary color images, allowing for the reduction of brightness in similar parts and accurate recognition of targets without pre-setting exclusion RGB values.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If RGB values and exclusion RGB values are set in advance for image processing, then the inspection process can be performed, but the burden of setting processing becomes large and accurate extraction becomes difficult when lighting conditions or object orientation change

Engineering Contradiction:
Improveinspection process efficiencyVSAvoidsetting processing burden
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The system automatically determines exclusion RGB values by analyzing the captured image itself, without requiring manual pre-setting. The control device extracts the main color component from the captured image and automatically sets exclusion values based on complementary colors, allowing the system to serve itself rather than requiring operator intervention for parameter setting.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system dynamically adjusts RGB parameter settings based on the actual captured image data. Instead of using fixed pre-set values, the exclusion RGB values are changed automatically according to the main color components detected in the specific image being processed, adapting parameters to match actual lighting and orientation conditions.

Inventive Principle:
Principle #35Parameter changes

2Device complexity

If fixed RGB values are used for recognition, then the processing is simple, but accuracy decreases when lighting conditions or object orientation change causing RGB value variations

Engineering Contradiction:
Improveprocessing simplicityVSAvoidrecognition accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The system transitions from static, pre-set RGB values to dynamic parameter adjustment. The exclusion RGB values are determined automatically for each captured image based on its specific lighting and orientation conditions, allowing the recognition parameters to adapt dynamically rather than remaining fixed.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system uses feedback from the captured image itself to determine appropriate exclusion RGB values. By analyzing the main color components in the actual captured image and using complementary color theory, the system automatically sets exclusion values that are appropriate for the specific lighting and orientation conditions observed in the feedback from the image sensor.

Inventive Principle:
Principle #23Feedback

3Reliability

If manual setting of exclusion RGB values is performed, then specific targets can be excluded, but it is difficult to accurately extract RGB values when lighting or orientation changes

Engineering Contradiction:
Improvetarget exclusion capabilityVSAvoidparameter setting difficulty
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The system automatically determines exclusion RGB values by analyzing the captured image itself, without requiring manual pre-setting. The control device extracts the main color component from the captured image and automatically sets exclusion values based on complementary colors, allowing the system to serve itself rather than requiring operator intervention for parameter setting.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP4056992B1Image processing for post-mounting inspection of PCB parts
Publication Date: 2025.02.19 FUJI CORP
  • EP4056992B1 patent drawingFigure 1
  • EP4056992B1 patent drawingFigure 2
  • EP4056992B1 patent drawingFigure 3

AI summary

An image processing device that processes a color image in which each pixel has gradation values of three primary colors of RGB, includes: an image acquiring section configured to acquire an image including a recognition target and a similar part of which a main color component is different from a main color component of the recognition target and brightness is similar to brightness of the recognition target, as the color image; a difference image generating section configured to use a first primary color image in which a gradation value of a first primary color that is close to the main color component of the similar part, out of the three primary colors of RGB, is extracted from the color image and a second primary color image in which a gradation value of a second primary color except the first primary color is extracted from the color image, to generate a difference image having a gradation value based on a difference obtained by subtracting the gradation value of the second primary color image from the gradation value of the first primary color image; a recognition image generating section configured to generate a recognition image having a gradation value obtained by subtracting the gradation value of the difference image from a gradation value of an image in which any one of the three primary colors of RGB is extracted from the color image; and a recognition processing section configured to perform recognition processing of the recognition target using the recognition image.