Image Processing Apparatus Defect Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image processing methods struggle to effectively extract defects from printed images without also identifying unique portions, leading to either missing defects in solid areas or incorrectly identifying non-defects as unique portions.

Innovation Solution

An image processing apparatus and method that employs a unique portion detecting algorithm, which involves generating inspection target image data, setting inspection target areas, and performing a unique portion detecting process by averaging and quantizing image data across various division sizes and phases, while suppressing the effect of input image data to focus on specific defects like white stripes caused by ejection failures.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a unique portion detecting algorithm is applied to extract defects from printed images, then defect detection capability is improved, but unique portions of the printed image are also simultaneously detected as false positives

Engineering Contradiction:
Improvedefect detection accuracyVSAvoidfalse positive rate
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The image is divided into multiple division areas, and processing is performed independently on each area. This segmentation allows the system to distinguish between local defect patterns and global unique portions, reducing false positives while maintaining defect detection accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the parameter of division area size and phase to process the same image at multiple scales and orientations. By averaging results across multiple parameters, the system can differentiate between consistent defect patterns and variable unique portions, improving reliability.

Inventive Principle:
Principle #35Parameter changes

2Reliability

If only solid areas are inspected to avoid false positives, then reliability is improved, but defects in non-solid areas cannot be extracted

Engineering Contradiction:
Improvefalse positive reductionVSAvoiddefect detection coverage
Core Design Contradiction:
ReliabilityVSAdaptability or versatility

Solution Approach 1:

The inspection system is designed to handle multiple types of areas (solid, patterned, gradient) using the same unique portion detecting algorithm. This universal approach allows the system to maintain high reliability across all area types while expanding defect detection coverage to include previously missed defects in non-solid areas.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Adaptability or versatility

If the entire image is processed to detect all possible defects, then defect detection coverage is improved, but processing load and time increase

Engineering Contradiction:
Improvedefect detection coverageVSAvoidinspection speed
Core Design Contradiction:
Adaptability or versatilityVSProductivity

Solution Approach 1:

By segmenting the image into division areas and processing them independently with parallel computation, the system achieves full image coverage while reducing the computational burden on any single processing unit, thereby maintaining high inspection speed.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies processing with multiple division sizes and phases (excessive action) but averages the results to achieve the same effect as a single comprehensive processing pass, reducing overall computation time while maintaining detection coverage.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3113473B1Image processing apparatus and image processing method
Publication Date: 2020.01.29 CANON KK
  • EP3113473B1 patent drawingFigure 1A~1D
  • EP3113473B1 patent drawingFigure 2
  • EP3113473B1 patent drawingFigure 3

AI summary

An image processing apparatus has: a generating unit configured to generate inspection target image data on the basis of read image data; a setting unit configured to set inspection target areas from the inspection target image data; and an extracting unit configured to extract a unique portion by applying a predetermined process to the set inspection target areas. The setting unit sets the inspection target areas so as to make the ratio of inspection target areas to the entire image area of the inspection target image data in a predetermined direction larger than the ratio of inspection target areas to the entire image area in a direction intersecting with the predetermined direction.