Vision Inspection Focus Determination Using Single Image Feature Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining the in-focus position in image analysis, such as those used in computer-aided biological material inspection, require capturing multiple images and frequent system adjustments, leading to inefficiencies and potential wear and tear on the vision inspection system.
Innovation Solution
A method that determines the difference between a sample position and an in-focus position using image data, allowing for the capture of fewer images, often just one, by employing an image processing device with a processor to analyze contrast, directional, and content features extracted from the image data, and adjusting the optical system accordingly to achieve focus.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple images are captured at different positions to determine in-focus position, then measurement precision is improved, but loss of time increases and device complexity increases
Solution Approach 1:
The patent extracts only the necessary information (focus quality indicators) from a single image to determine in-focus position, rather than capturing multiple images. The image processing device extracts focus quality indicators from captured image data to identify the in-focus position, reducing the number of images needed while maintaining measurement precision.
Solution Approach 2:
The patent replaces the mechanical approach of capturing multiple images at different positions with an information processing approach. Instead of physically moving the imaging system multiple times, the system processes image data to extract focus quality indicators and determines the in-focus position computationally, significantly reducing processing time.
2Measurement precision
If multiple images are captured at different positions to determine in-focus position, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent extracts focus quality indicators from a single image to determine the in-focus position, eliminating the need for multiple image captures and system readjustments. The image processing device extracts the necessary focus information directly from the captured image data, reducing device complexity and readjustment frequency.
Solution Approach 2:
The imaging system performs self-diagnosis of focus quality by extracting focus quality indicators from its own captured image data. The system automatically determines whether the captured image is in-focus or out-of-focus without requiring external intervention or multiple readjustment cycles, reducing overall system complexity.
3Measurement precision
If frequent system readjustments are performed to capture images at different positions, then measurement precision is improved, but reliability decreases due to wear and tear
Solution Approach 1:
The patent extracts focus quality indicators from a single image to determine the in-focus position, eliminating the need for frequent system readjustments and image recaptures. This reduces mechanical wear and tear on the imaging system, improving reliability and durability while maintaining measurement precision.
4Productivity
If a single image is used to determine in-focus position, then productivity is improved, but measurement precision may deteriorate
Solution Approach 1:
The patent changes the approach from capturing multiple images at different positions to analyzing a single image with multiple parameters. The image processing device extracts multiple focus quality indicators (such as sharpness, contrast, and edge definition) from the single image data to accurately determine the in-focus position, maintaining measurement precision while improving productivity.
Data Source
Figure 1
Figure 2
Figure 3~5
AI summary
The present invention relates to a method for determining a difference (Delta Z) between a sample position (Zobj) and an in-focus position (Zf), as well as a vision inspection system. In a first step image data depicting a sample (304a, 304b) is captured. Next, a feature set is extracted from the image data. Thereafter, the feature set is classified into a position difference value, corresponding to the difference between the sample position and the in-focus position, by using a machine learning algorithm that is trained to associate image data features to a position difference value.