Screening Method for Object Detection Using Image Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current screening methods for detecting objects of interest, such as asbestos fibers, are hindered by the complex surface structure of sample carriers like gold-coated filters, leading to increased operator fatigue and reduced detection efficiency due to vigilant tasks associated with interpreting complex images.
Innovation Solution
The method involves acquiring images of both the sample carrier and the sample, using image processing techniques like the Hough transform or neural networks to manipulate the images, reducing the surface structure's prominence, thereby simplifying the detection of objects of interest using a charged particle microscope with detectors like backscatter and secondary electron detectors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If images are acquired using a charged particle microscope to detect objects of interest, then detection capability is provided, but the complex surface structure of the sample carrier obscures objects and increases image complexity
Solution Approach 1:
The patent extracts and removes the sample carrier surface structure from the acquired image through image processing techniques, isolating only the objects of interest. This extraction process eliminates the obscuring effect of the complex surface structure while preserving the detection capability of the charged particle microscope.
Solution Approach 2:
The patent introduces image processing as an intermediary step between image acquisition and object detection. This intermediary process manipulates the image data to reduce complexity from the sample carrier surface structure, making the subsequent detection task easier without compromising the original detection capability.
2Productivity
If human operators screen samples for detecting objects of interest, then detection is performed, but operator fatigue increases and detection accuracy decreases over time
Solution Approach 1:
The patent replaces the mechanical human visual inspection process with automated image processing and analysis systems. This substitution eliminates operator fatigue and vigilance decrement while maintaining or improving detection accuracy through consistent, repeatable automated analysis of the manipulated images.
3Loss of information
If the surface structure of the sample carrier is visible in acquired images, then sample carrier information is preserved, but object detection becomes more difficult due to obscured features
Solution Approach 1:
The patent extracts and removes the sample carrier surface structure information from the acquired images through image processing, isolating only the objects of interest. This extraction eliminates the obscuring effect of the surface structure while preserving all relevant object information for detection.
Solution Approach 2:
The patent segments the acquired image into separate components: sample carrier surface structure and objects of interest. By separating these elements, the processing system can eliminate the surface structure component while preserving and enhancing the visibility of the objects of interest for easier detection.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the ease and accuracy of object detection by reducing image complexity, allowing human operators to more effectively screen samples with reduced fatigue and improving the overall efficiency of the screening process.
Implementation Method 1
acquiring an image of said sample using a charged particle microscope
Implementation Method 2
a backscatter electron detector device (22) that is arranged for acquiring said image of said sample
Implementation Method 3
a secondary electron detector device (22) that is arranged for acquiring said image of said sample
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a screening method. The method comprises the step of providing a sample, wherein said sample comprises a sample carrier with a surface structure, as well as an object of interest. The method further comprises the step of acquiring an image of said sample. According to the disclosure, the method comprises the steps of providing information on said surface structure of said sample carrier, which may in particular comprise the step of acquiring an image of said sample carrier. In that case two images are obtained: one more sensitive to the objects of interest, and one more sensitive to the surface structure of the sample carrier. This allows manipulation of the acquired image, using said information on the surface structure of the sample carrier. With this, said manipulated image may be screened for easy and reliable detection of said object of interest.