Semiconductor spray head point position derivation detection method based on boundary tracking algorithm
Through an automated detection method based on the boundary tracking algorithm, the problem of degradation of detection accuracy caused by blockage of holes in semiconductor spray heads and foreign matter residues is solved, efficient and accurate point detection is achieved, and production efficiency and quality control are improved.
Patent Information
- Application Number
- CN202510230560.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-28
- Publication Date
- 2025-06-06
AI Technical Summary
The blockage of the hole position and foreign matter residue of the semiconductor spray head leads to a decrease in detection accuracy, and the blocked hole position cannot be accurately identified and positioned, affecting quality control and production efficiency.
The point derivation detection method based on the boundary tracking algorithm is adopted to initialize the basic samples through automated means, identify the reference image information, correct the image to be tested, and mark the point contour area to achieve high-precision detection.
It improves the automation degree and detection accuracy of semiconductor spray head point detection, reduces manual intervention, and ensures the accuracy and reliability of detection results.
Smart Images
Figure CN120107221A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of image recognition technology related to semiconductor manufacturing, and in particular to a semiconductor shower head point derivation detection method based on a boundary tracking algorithm. Background Art
[0002] In the current semiconductor shower head (ShowerHead), although there are many circular holes distributed on its surface to meet specific process requirements, these holes have exposed a series of problems after long-term use. Specifically, hole blockage and foreign matter residue have become defects that cannot be ignored, which have a significant impact on the performance of the semiconductor shower head (ShowerHead) and the subsequent inspection process.
[0003] First, hole blockage and foreign matter residue will cause some circular holes to gradually deviate from their original circular pattern. This deformation not only affects the uniformity of fluid distribution in the semiconductor shower head (ShowerHead), but also poses a challenge to the subsequent boundary tracking algorithm. As an advanced technology based on image processing, the accuracy of the boundary tracking algorithm is highly dependent on the shape and contour of the object being detected. When the circular hole is deformed, the recognition accuracy of the algorithm will drop significantly, and even misjudgment or missed judgment may occur.
[0004] Secondly, existing inspection cameras cannot detect completely blocked round holes when taking photos. This means that these blocked holes will be completely ignored in the subsequent image recognition process, further exacerbating the problem of underestimation of defects. Since these blocked holes cannot be accurately identified and located, quality control in the production process will be severely affected, which may lead to the outflow of substandard products.
[0005] Therefore, developing an efficient and accurate point detection method is of great significance to improving product quality and production efficiency in the semiconductor manufacturing process. Summary of the invention
[0006] The purpose of the present invention is to overcome the deficiencies of the prior art and provide a semiconductor shower head point derivation detection method based on a boundary tracking algorithm, which achieves accurate point detection through automated means, thereby improving detection efficiency and accuracy.
[0007] In order to achieve the above-mentioned purpose, the present invention provides a semiconductor shower head point derivation detection method based on a boundary tracking algorithm, which includes the following steps: S1. Initialize the pre-input basic samples and obtain the reference image information based on the boundary tracking algorithm; S2. Input the image to be inspected, and obtain the inspection image information based on the boundary tracking algorithm; S3. According to the reference image information and the detection image information, the angle offset of the image to be tested is obtained by comparison with the base sample, and the image to be tested is corrected based on the angle offset; S4. Based on the boundary tracking algorithm, randomly select a central pixel point of a test circle from the image to be tested as the starting point; S5. Determine the center pixel points of each adjacent inspection point around the starting point based on the reference adjacent angle and the distance between the reference points in the reference image information, and calculate and mark the contour area of each adjacent inspection point based on the reference point diameter width in the reference image information; S6. Taking the center pixel point of the neighboring inspection dots obtained last time as the starting point, repeat steps S5-S6 cyclically to determine a new starting point until there are no new inspection dots that can be marked in the image to be inspected.
[0008] Further, in step S1, the basic sample is initialized to determine the number of adjacent reference dots surrounding a reference dot.
[0009] Further, in step S1, the reference image information includes the center pixel coordinate position of each reference point in the basic sample, the reference point diameter, the distance between the center pixels of two adjacent reference points, and the reference adjacent angle between the center pixels of two adjacent reference points.
[0010] Further, in step S2, the inspection image information includes the central pixel coordinate position of each inspection dot in the basic sample, the diameter of the inspection dot, the distance between the central pixels of two adjacent inspection dots, and the inspection adjacent angle between the central pixels of two adjacent inspection dots.
[0011] Further, in step S3, the angle offset is the difference between the reference adjacent angle and the verification adjacent angle.
[0012] Furthermore, in step S5, the distance between the reference points is an average value of distances between the center pixels of any two adjacent reference points calculated by multiple accumulations.
[0013] Further, in step S5, the marked inspection dot contour area is compared with the basic sample points.
[0014] The present invention adopts the above-mentioned solution, and its beneficial effects are: 1) The automation level of semiconductor sprinkler head point detection is improved, reducing the need for manual intervention.
[0015] 2) High-precision point detection is achieved through boundary tracking algorithm and precise parameter calculation.
[0016] 3) Provides result verification and optimization steps to ensure the accuracy and reliability of the test results. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Figure 1 Schematic diagram of the process for deriving the detection method. DETAILED DESCRIPTION
[0018] In order to facilitate the understanding of the present invention, the present invention is described more fully below with reference to the accompanying drawings. The accompanying drawings provide preferred embodiments of the present invention. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. The purpose of providing these embodiments is to make the disclosure of the present invention more thoroughly and comprehensively understood.
[0019] See attached Figure 1 As shown, in this embodiment, a semiconductor shower head point derivation detection method based on a boundary tracking algorithm is provided. The method realizes accurate detection and judgment of points by automated means, thereby deducing and marking all points, thereby improving detection efficiency and accuracy. Specifically, the derivation detection method of the present invention comprises the following steps: Step S1. Initialize the pre-input basic sample, and obtain the reference image information based on the boundary tracking algorithm, wherein, in step S1, the basic sample is initialized to confirm the number of adjacent reference points arranged around a reference point, and secondly, in step S1, the reference image information includes the central pixel coordinate position of each reference point in the basic sample, the reference point diameter, the distance between the central pixels of two adjacent reference points, and the reference adjacent angle between the central pixels of two adjacent reference points. The reference adjacent angle can be confirmed according to the number of adjacent reference points established in the initialized basic sample, for example: combined with the common semiconductor shower head specifications, generally one reference point is the center, and six reference points are arranged around it, and accordingly, the reference adjacent angle is 360° / 66=60°, and the corresponding distribution orientations of the six adjacent reference points are 0° (360°), 60°, 120°, 180°, 240°, and 300°.
[0020] Step S2. Input the image to be inspected, and obtain the inspection image information based on the boundary tracing algorithm. In step S2, the inspection image information includes the coordinate position of the center pixel point of each inspection point in the basic sample, the diameter of the inspection point, the distance between the center pixel points of two adjacent inspection points, and the inspection adjacent angle between the center pixel points of two adjacent inspection points.
[0021] Step S3. According to the reference image information and the detection image information, the angle offset of the image to be tested compared with the basic sample is obtained, and the image to be tested is corrected based on the angle offset. Specifically, in step S3, the angle offset is the difference between the reference adjacent angle and the detection adjacent angle. For ease of understanding, the following is explained in conjunction with relevant examples: when the detection adjacent angle is calculated to be 66°, the reference adjacent angle that matches it is 60°, and the difference between the two is 6°. This difference is used as the angle that the image to be tested needs to be rotated, so as to rotate and correct the image to be tested. Therefore, the image to be tested is corrected by rotation, so that the angle orientation of the sample to be tested and the basic sample are matched and consistent, which is more conducive to subsequent point inference.
[0022] Step S4. Based on the boundary tracking algorithm, randomly select the central pixel of a test circle from the image to be tested as the starting point Step S5. Based on the reference adjacent angle and the distance between reference points in the reference image information, the center pixel point of each adjacent neighboring inspection circle set around the starting point is determined in sequence based on the boundary tracing algorithm, with the starting point as the center, according to the reference adjacent angle and the distance between reference points, and further explanation is given with a specific example: first, for example, the coordinates of the starting point are (0, 0), the reference adjacent angle is 60°, and the distance between reference points is , thus it can be concluded that the central pixels of the six neighboring test points are ( ,0),(2,1),(-2,1),(- , 0), (-2, -1), (2, -1). Secondly, the contour area of each adjacent inspection dot is calculated and marked based on the reference dot diameter width in the reference image information, wherein the contour area of the inspection dot is the (Xmin, Ymax) and (Xmax, Ymin) areas that can cover the current center pixel point calculated by the reference dot diameter width obtained from the reference image information.
[0023] Step S6. Taking the center pixel point of the neighboring inspection dots obtained last time as the starting point, repeat steps S5-S6 cyclically to determine a new starting point until there are no new inspection dots that can be marked in the image to be inspected.
[0024] In this embodiment, in step S5, the accuracy of the marking result is verified by comparing the marked test dot contour area with the basic sample points. When the points are consistent, the marking is considered successful and no additional operation is performed; otherwise, when the points are inconsistent, the process returns to step S5 for re-comparison until a satisfactory marking accuracy is achieved.
[0025] In summary, the above steps can realize the standardized processing, correction and point inference marking of the image to be inspected, providing an accurate and reliable data basis for subsequent image analysis.
[0026] In this embodiment, in step S5, the distance between reference points is an average value of distances between center pixels of any two adjacent reference points calculated multiple times.
[0027] In addition, the boundary tracking algorithm described in this embodiment is an effective image processing technology that can identify boundaries in an image and track these boundaries, thereby accurately locating points in the image. The algorithm has important application value in semiconductor shower head point detection because it can handle complex image backgrounds and maintain a high recognition accuracy rate under incomplete identification conditions of semiconductor shower head points.
[0028] In practical applications, this detection method can be integrated into an automated detection system, and the above steps can be automatically performed through a software program, thereby achieving fast and accurate semiconductor shower head point detection. This not only improves production efficiency, but also reduces the errors that may be introduced by manual detection, ensuring quality control of the semiconductor manufacturing process.
[0029] In summary, the present invention provides an efficient and accurate method for detecting the position of a semiconductor shower head, which realizes accurate detection and marking of the position of a semiconductor shower head through automated means, and provides powerful technical support for quality control in the semiconductor manufacturing process.
[0030] The embodiments described above are only preferred embodiments of the present invention and are not intended to limit the present invention in any form. Any technician familiar with the art who, without departing from the scope of the technical solution of the present invention, makes more possible changes and modifications to the technical solution of the present invention using the technical content disclosed above, or modifications are all equivalent embodiments of the present invention. Therefore, any equivalent and equivalent changes made according to the ideas of the present invention without departing from the content of the technical solution of the present invention should be included in the protection scope of the present invention.
Claims
1. A semiconductor shower head point derivation detection method based on a boundary tracking algorithm, characterized in that: The following steps are included: S1. Initialize the pre-input basic samples and obtain the reference image information based on the boundary tracking algorithm; S2. Input the image to be inspected, and obtain the inspection image information based on the boundary tracking algorithm; S3. According to the reference image information and the detection image information, the angle offset of the image to be tested is obtained by comparison with the base sample, and the image to be tested is corrected based on the angle offset; S4. Based on the boundary tracking algorithm, randomly select a central pixel point of a test circle from the image to be tested as the starting point; S5. Determine the center pixel points of each adjacent inspection point around the starting point based on the reference adjacent angle and the distance between the reference points in the reference image information, and calculate and mark the contour area of each adjacent inspection point based on the reference point diameter width in the reference image information; S6. Taking the center pixel point of the neighboring inspection dots obtained last time as the starting point, repeat steps S5-S6 cyclically to determine a new starting point until there are no new inspection dots that can be marked in the image to be inspected.
2. According to claim 1, a semiconductor shower head point derivation detection method based on boundary tracking algorithm is characterized in that: In step S1, a basic sample is initialized to determine the number of adjacent reference dots surrounding a reference dot.
3. The semiconductor shower head point derivation detection method based on boundary tracking algorithm according to claim 1 is characterized in that: In step S1, the reference image information includes the center pixel coordinate position of each reference dot in the basic sample, the reference dot diameter, the distance between the center pixels of two adjacent reference dots, and the reference adjacent angle between the center pixels of two adjacent reference dots.
4. The semiconductor shower head point derivation detection method based on boundary tracking algorithm according to claim 3 is characterized in that: In step S2, the inspection image information includes the central pixel coordinate position of each inspection dot in the basic sample, the diameter of the inspection dot, the distance between the central pixels of two adjacent inspection dots, and the inspection adjacent angle between the central pixels of two adjacent inspection dots.
5. The semiconductor shower head point derivation detection method based on boundary tracking algorithm according to claim 4 is characterized in that: In step S3, the angle offset is the difference between the reference adjacent angle and the verification adjacent angle.
6. The semiconductor shower head point derivation detection method based on boundary tracking algorithm according to claim 3 is characterized in that: In step S5, the distance between the reference points is an average value of distances between the center pixels of any two adjacent reference points calculated multiple times.
7. The semiconductor shower head point derivation detection method based on boundary tracking algorithm according to claim 3 is characterized in that: In step S5, the marked inspection dot contour area is compared with the basic sample points.