X-ray Inspection Apparatus Subpixel Deviation Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional X-ray inspection apparatuses face limitations in achieving high geometric magnification and resolution due to physical constraints such as focal point size and X-ray dose, leading to image noise and elongated photographing times, and are costly due to the need for high-positioning accuracy components.
Innovation Solution
An X-ray inspection apparatus with a movement mechanism that adjusts the relative positions of the X-ray source, detector, and object to acquire multiple images with subpixel positional deviations, utilizing repeated positioning errors for super-resolution reconstruction processing without requiring high-accuracy mechanical components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the distance from the X-ray source to the object is shortened to increase geometric magnification, then image resolution is improved, but blurring increases due to the finite focal point size
Solution Approach 1:
The patent applies dynamics by moving the X-ray detector to multiple different positions relative to the object during image acquisition. The detector is positioned at several locations around the object, and images are captured from each position. This dynamic positioning allows the system to overcome the blurring effect caused by finite focal point size while maintaining high geometric magnification, as each detector position captures slightly different projection data that can be combined to reduce blurring artifacts.
Solution Approach 2:
The patent transitions from a single static imaging geometry to a multi-position imaging geometry. Instead of relying solely on optimizing the source-to-object distance in one dimension, the system introduces the detector position as an additional degree of freedom. By capturing images from multiple detector positions around the object, the system effectively adds spatial dimensionality to the data acquisition process, enabling super-resolution reconstruction that overcomes the blurring limitation.
2Measurement precision
If the focal point size is reduced to decrease blurring, then image resolution is improved, but X-ray dose decreases leading to increased image noise
Solution Approach 1:
The patent merges multiple low-dose images acquired from different detector positions to produce a single high-resolution image with reduced noise. Instead of using a single high-magnification image that would require a small focal point and result in low X-ray dose per pixel, the system combines information from multiple images taken at different detector positions. This merging process allows the use of a larger focal point (higher dose per image) while achieving superior resolution and lower noise through the combination of multiple measurements.
Solution Approach 2:
The patent implements continuous useful action by acquiring multiple images in sequence from different detector positions without moving the object. The X-ray source continuously irradiates the object while the detector is positioned at various locations, and all these measurements are integrated to form the final high-resolution image. This continuous acquisition process maximizes the use of X-ray dose by ensuring that each photon contributes to the final reconstructed image, thereby reducing noise while maintaining high resolution.
3Measurement precision
If an X-ray detector with small pixels and large number of pixels is used to improve resolution, then image resolution is improved, but detector size decreases narrowing the visual field
Solution Approach 1:
The patent uses dynamic positioning of the detector to overcome the limited visual field of small detectors. Instead of relying on a single large detector, the system moves the detector to multiple positions around the object, effectively scanning a larger volume. This dynamic approach allows a small detector with high pixel density to capture images from multiple angles, and the combined data provides both high resolution and a wide effective visual field that would require a much larger single detector.
4Measurement precision
If the distance between the X-ray source and the X-ray detector is increased to increase geometric magnification, then image resolution is improved, but the apparatus size increases and X-ray dose entering the detector decreases
Solution Approach 1:
The patent solves the apparatus size problem by changing from a linear arrangement (source-object-detector in a straight line with large source-to-detector distance) to a three-dimensional arrangement where the detector is positioned at multiple locations around the object. This spatial reconfiguration allows high geometric magnification to be achieved without increasing the overall apparatus footprint, as the detector can be placed close to the object from different angular positions rather than requiring a large linear distance.
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
Enables the generation of high-resolution images without the need for expensive high-accuracy movement mechanisms, reducing costs and photographing time while overcoming limitations in geometric magnification and resolution.
Implementation Method 1
an X-ray source 41 composed an X-ray tube and configured to irradiate a workpiece W, which is an object, with X rays from below to above, and an X-ray detector 42 configured to detect X-rays transmitted through the workpiece W after the irradiation from the X-ray source 41
Data Source
AI summary
Utilizing random variation (repeated positioning error) when reciprocating operation is repeatedly performed in which a stage is moved by (+x, +y) pulses toward an arbitrary position perpendicular to an optical axis of X-rays extending from an X-ray source to an X-ray detector, and then, is moved from there by (−x, −y) pulses, an image group of images obtained by moving in parallel to each other is acquired, and an image processing unit finds a deviation between the images, and acquires an input image group in which each of the images has the deviation at a subpixel level. The image processing unit executes a reconstruction processing, using the input image group in which each of the images has the deviation at the subpixel level to generate a super-resolution image.