Image Inspecting Apparatus Multi-Threshold Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image inspecting apparatuses require varying threshold values for accurate detection, making it difficult to obtain desired detection results in a single image inspection, leading to suboptimal performance.
Innovation Solution
An image inspecting apparatus and program that automatically compare reference-image and inspection-image data by establishing correspondences at a pixel level, performing image matching, detecting differences, and producing difference point representing image data using multiple threshold values for enhanced detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single threshold value is used for image inspection, then the inspection process is simple and fast, but the detection accuracy is insufficient
Solution Approach 1:
The patent divides the image inspection process into multiple independent inspection regions, each with its own threshold value. This segmentation allows different threshold values to be applied to different regions simultaneously, improving detection accuracy without requiring a completely complex re-inspection process. The system maintains simplicity by processing all regions in parallel within a single inspection cycle.
Solution Approach 2:
The patent implements dynamic threshold adjustment where the threshold value is automatically varied based on the inspection region and detected difference characteristics. The system dynamically selects appropriate threshold values from multiple predetermined levels for each region, enabling high detection accuracy while maintaining a streamlined single-pass inspection process.
2Measurement precision
If the threshold value is varied manually to improve detection accuracy, then detection results can be optimized, but multiple inspection cycles are required and time is lost
Solution Approach 1:
The patent performs preliminary preparation by storing multiple predetermined threshold values in advance and automatically selecting appropriate thresholds for different inspection regions before the actual inspection begins. This preliminary configuration eliminates the need for manual threshold adjustment during inspection cycles, achieving optimized detection results in a single pass without time loss.
Solution Approach 2:
The inspection system performs self-adjustment by automatically varying threshold values based on the characteristics of each inspection region and detected differences. The system serves itself by selecting and applying optimal thresholds without external manual intervention, achieving both high detection accuracy and efficient single-pass inspection timing.
3Measurement precision
If multiple threshold values are applied to different regions, then detection accuracy improves, but the system complexity increases
Solution Approach 1:
The patent segments the inspection image into multiple independent regions, each associated with specific threshold values stored in advance. This segmentation approach allows multiple thresholds to be managed systematically without overwhelming complexity, as each region processes independently with its predetermined threshold set.
Solution Approach 2:
The patent uses predetermined threshold values that are copied and stored in advance for different inspection regions. Instead of calculating or adjusting thresholds in real-time, the system retrieves and applies pre-configured threshold copies, reducing computational complexity while maintaining high detection accuracy across multiple regions.
Data Source
AI summary
Image inspecting apparatus compares first image data created as data representing a reference-image acting as an inspecting reference with second image data created as data representing an inspection-image acting as a target to automatically extract a difference point between first and second image data, and includes a storage means for the reference- and inspection-image, an image processing means for establishing correspondences between part of stored reference-image as first image data with part of stored target image as second image data at a pixel level to perform an image matching processing of them, a difference detecting means for comparing image-matched first and second image data to detect a difference between first and second image data, image producing means for comparing difference with a plurality of threshold values to produce error representing image data at each threshold value, and inspecting process using produced error representing image data at each threshold value.