Method and Measurement System for Warpage Detection of Electronic Packaging Substrate Combined with Super-Resolution Reconstruction

Abstract

The present invention belongs to the technical field related to the warpage detection of microelectronic substrates, and discloses a method and a measurement system for detecting the warpage of an electronic packaging substrate combined with super-resolution reconstruction. The method includes the following steps: S1 Set the speckle diameter of the circular speckle pattern, project the circular speckle pattern on the surface of the test sample, collect multiple pictures of the projected pattern, draw a two-dimensional graph of the gray value and the corresponding ratio P of each gray value, fit the two-dimensional graph into a bimodal distribution pattern and obtain the function of the ratio P, that is, the prior function; S2 Replace the test sample with the sample to be measured and collect multiple photos. Use the prior function to perform super-resolution reconstruction on the collected multiple photos according to a preset magnification factor to obtain a high-resolution image; S3 Use the obtained high-resolution image to calculate the three-dimensional topography of the surface of the sample to be measured, so as to obtain the warpage deformation information of the sample to be measured. Through the present invention, the problem of low detection accuracy of substrate warpage deformation is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field related to the warpage detection of microelectronic substrates, and more specifically, relates to a method and a measurement system for detecting the warpage of an electronic packaging substrate by combining super-resolution reconstruction. Background Art

[0002] Compared with traditional contact detection methods, such as vernier calipers and micrometer measurements, non-contact measurement methods are widely used in the deformation detection of precision devices, such as the warpage of electronic packaging substrates and the deformation detection of materials, because they do not affect the surface characteristics of the test specimens. Currently, widely used non-contact measurement methods include the Three-dimensional Digital Image Correlation (3D-DIC) method, the Moiré fringe method, etc. The 3D-DIC method is widely used in the technical field related to the warpage detection of microelectronic substrates due to its advantages of simple device and low cost.

[0003] When measuring by the 3D-DIC method, the surface of the test specimen needs to contain sufficient gray-scale change information, which is generally obtained by preparing speckles on its surface. The acquisition of speckles includes spraying speckles and projecting speckles. Due to the non-invasive nature of projecting speckles and the advantage of having no impact on the surface of the test specimen, it is widely used in the measurement of precision electronic packaging structures.

[0004] However, during the measurement process of projecting speckles by 3D-DIC, due to the influence of the vibration of the projector and the electronic noise in the projected pattern, the projected speckle pattern will contain tiny vibration displacements and the shape and brightness changes of individual speckles. When these changes are optically amplified, they will have an obvious impact on the measurement results, resulting in the failure of speckle pattern matching, reducing the success rate of subset matching, causing deviations in the measurement results, and reducing the accuracy of 3D-DIC measurement. Summary of the Invention

[0005] In view of the above defects or improvement requirements of the prior art, the present invention provides a method and a measurement system for detecting the warpage of an electronic packaging substrate by combining super-resolution reconstruction, so as to solve the problem of low accuracy in detecting the warpage deformation of the substrate.

[0006] To achieve the above object, according to one aspect of the present invention, a method for detecting the warpage of an electronic packaging substrate by combining super-resolution reconstruction is provided. The method includes the following steps:

[0007] S1 Set the speckle diameter of the circular speckle pattern, project the circular speckle pattern onto the surface of the test sample, collect multiple pictures of the projected pattern on the surface of the test sample, count the gray values of each pixel point in all the pictures, thereby obtaining the proportion P of the number of pixel points corresponding to each gray value to the total number of pixel points, draw a two-dimensional graph of the gray value and the proportion P corresponding to each gray value, and fit it into a bimodal distribution pattern according to this two-dimensional graph to obtain the function of the proportion P, that is, the prior function;

[0008] S2 Replace the test sample in step S1 with the sample to be measured, collect multiple photos of the projected pattern on the surface of the sample to be measured and obtain the gray values of all pixel points on each picture, and use the prior function in step S1 to perform super-resolution reconstruction on the multiple collected photos according to a preset magnification factor, thereby obtaining a high-resolution image;

[0009] S3 Calculate the three-dimensional topography of the surface of the sample to be measured by using the high-resolution image obtained in step S3, thereby obtaining the warpage deformation information of the sample to be measured.

[0010] Further preferably, in step S2, the prior function is in accordance with the following relational expression:

[0011]

[0012] wherein, μ0 and μ1 are respectively the gray values at the two aggregation peaks, σ0 and σ1 are respectively the standard deviations of the two peaks, H i,j is the pixel value at the i-th row and j-th column in the predicted reconstructed high-resolution image H, c is a normalization constant, P DIC [H i,j is the prior distribution function.

[0013] Further preferably, in step S2, the reconstruction adopts the maximum a posteriori probability algorithm, and the posterior probability function is in accordance with the following relational expression:

[0014]

[0015] wherein, L k is the low-resolution image sequence before reconstruction, H is the high-resolution image, P DIC [H i,j is the prior distribution function, and H i,j is the pixel value at the i-th row and j-th column in the predicted reconstructed high-resolution image H.

[0016] Further preferably, in steps S1 and S2, the photos of the test sample and the sample to be measured are both images collected from the left and right sides of the test sample or the sample to be measured respectively. In step S2, when reconstructing the high-resolution image, the images collected from the left and right sides are reconstructed separately, thereby obtaining the corresponding high-resolution images respectively.

[0017] Further preferably, in step S2, the preset magnification factor enables the reconstructed image to contain 3 to 5 clear speckle spots in a square subset with a side length of 33 pixels.

[0018] Further preferably, in step S2, the high-resolution image means that the length and width resolution of the reconstructed image is between 1.5 times and 2.5 times that of the image before reconstruction.

[0019] Further preferably, in step S1, the speckle pattern is white spots on a black background, and the spot shape is circular.

[0020] Further preferably, in step S3, the 3D morphology of the surface of the sample to be measured is calculated using the 3D-DIC method.

[0021] According to another aspect of the present invention, there is provided a measurement system for detecting the warping of an electronic packaging substrate combined with super-resolution reconstruction. The measurement system includes an image acquisition module, a prior function construction module, a super-resolution reconstruction module, and a three-dimensional morphology construction module, wherein,

[0022] The image acquisition module is used to acquire pictures of the circular speckle pattern projected on the test sample or the sample to be measured;

[0023] The prior function construction module is used to construct a prior function according to the gray values of the pixel points in the acquired pictures;

[0024] The super-resolution reconstruction module is used to reconstruct the acquired pictures into high-resolution images by using the super-resolution reconstruction method according to the prior function and the preset magnification factor;

[0025] The three-dimensional morphology construction module is used to obtain the three-dimensional morphology of the surface of the sample to be measured by using the high-resolution image.

[0026] Further preferably, the measurement system further includes a control center. The image acquisition module includes a projector, a binocular camera, and an optical experimental platform. The projector and the binocular camera are arranged above the optical experimental platform. The two cameras in the binocular camera are respectively distributed on both sides of the projector. The test sample or the sample to be measured is placed on the optical experimental platform, and the projector projects the circular speckle pattern on the test sample or the sample to be measured.

[0027] Generally speaking, compared with the prior art by the above technical solutions conceived by the present invention, the following beneficial effects are achieved:

[0028] 1. The present invention reconstructs the acquired images through super-resolution reconstruction technology (SR technology) to improve the image resolution, enabling the DIC analysis system to identify smaller speckles, divide more subsets, and improve the accuracy of the DIC analysis system without increasing the hardware cost of the measurement system;

[0029] 2. The present invention reconstructs multiple images into one image through SR technology, enhancing the information that appears multiple times in a set of images. Since electronic noise appears randomly and each image in a set is acquired at a certain time interval (usually 0.5 s), the noise points formed by electronic noise at specific positions do not appear multiple times in a set of images, thereby weakening the noise information and reducing the influence of electronic noise on the measurement results;

[0030] 3. In the present invention, multiple images are aligned to the same high-resolution grid through SR technology and then reconstructed into one image, thereby reducing the influence caused by device vibration. In addition, binocular cameras are used to acquire photos from the left and right perspectives respectively, and the images obtained by the left and right cameras are respectively subjected to super-resolution reconstruction to obtain high-resolution images, so as to obtain a three-dimensional topography of the surface of the sample to be measured with higher accuracy;

[0031] 4. In the present invention, when measuring samples of different sizes or testing samples at different working distances of the camera, the average diameter of the speckles and the magnification factor of the SR technology can be adjusted to adapt to different working conditions, so that each subset in the finally acquired image can contain 3 - 5 clear speckle spots for DIC analysis and calculation, improving the measurement accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 is a flowchart of improving the measurement accuracy of projection speckle digital image correlation by using SR technology constructed according to the preferred embodiment of the present invention;

[0033] Figure 2 is a schematic diagram of a projection speckle digital image correlation topography measurement system constructed according to the preferred embodiment of the present invention;

[0034] Figure 3 is a schematic diagram of an algorithm flow for obtaining a prior function constructed according to the preferred embodiment of the present invention;

[0035] Figure 4 is a schematic diagram of an SR reconstruction algorithm flow based on maximum a posteriori probability constructed according to the preferred embodiment of the present invention;

[0036] Figure 5 is a schematic diagram of a speckle sample constructed according to the preferred embodiment of the present invention, where the speckle diameter is 3.0 pixels;

[0037] Figure 6It is a schematic diagram of the size of the stepped topography block used in the preferred embodiment of the present invention, where ROI represents the region of interest for detection. Among them, (a) is the sectional size diagram of the region of interest, and (b) is the schematic diagram of the three-dimensional model of the region of interest;

[0038] Figure 7 It is a diagram of the three-dimensional digital image correlation (3D-DIC) topography measurement results constructed according to the preferred embodiment of the present invention.

[0039] In all the drawings, the same reference numerals are used to represent the same structures, where:

[0040] 1 - optical experimental platform, 2 - test specimen, 3 - CCD camera, 4 - LCD projector, 5 - computer software analysis platform. Detailed implementation manners

[0041] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0042] As Figure 1 shown, a method for improving the accuracy of projection speckle digital image correlation deformation measurement includes the following steps:

[0043] S1. Prepare the test specimen and place it on the optical platform. According to the size of the specimen and the size of the region of interest, adjust the internal and external parameters of the dual CCD cameras and the parameters of the projector of the measurement system so that the projected speckle pattern can clearly cover the area to be measured of the test specimen and present a clear and high-contrast image in the computer acquisition system. After the adjustment is completed, calibrate the binocular cameras. The test specimen or the sample to be measured is a substrate.

[0044] S2. Adjust the size of the speckle diameter of the generated circular speckle pattern and the magnification of the SR reconstruction so that in the reconstructed image, within a square subset with a side length of 33 pixels, 3 - 5 clear speckle spots are included. In the subsequent measurement, the speckle diameter and the SR magnification remain unchanged.

[0045] S3. Collect multiple specimen pictures, export all the images and count the gray values of each pixel, and fit the proportion of the pixel points with each gray value into a bimodal distribution pattern as the prior distribution of the SR reconstruction to improve the accuracy and speed of the subsequent SR reconstruction.

[0046] S4. Start measuring the morphology of the specimen. The left and right cameras respectively collect a set of image sequences of the specimen to be tested, including seven images with a collection time interval of 0.5 s. Use the bimodal distribution probability density function fitted previously as the prior function, and reconstruct the image sequences collected by the left and right cameras into a high-resolution image, with the reconstruction magnification being the multiple determined in step 2.

[0047] S5. Take the reconstructed high-resolution image as the input for digital image correlation calculation, calculate the three-dimensional morphology results of the specimen surface, and thus obtain information such as the warping deformation of the specimen structure.

[0048] Further preferably, in step S1, both the projector and the binocular camera are set up on an optical vibration isolation platform to reduce the interference of external vibrations.

[0049] Further preferably, in S2, the speckle pattern generated by the computer is projected onto the specimen through the projector. By adjusting the projector focal length, the region of interest of the specimen is covered with a clear speckle pattern.

[0050] Further preferably, in S2, the style of the speckle pattern is white spots on a black background, the shape of the spots is circular, the average diameter of the speckles can be adjusted by computer software, the diameter unit is pixels, and the diameter range is generally from 2.0 to 4.0 pixels. Correspondingly, the magnification of SR reconstruction can also be set, generally set to 1.5 times, 2.0 times, and 2.5 times.

[0051] Further preferably, the calibration plate uses a dot calibration plate containing three concentric circles.

[0052] Further preferably, in step S3, as Figure 5 shown, when collecting the speckle images, the experimental parameters and environment remain unchanged, so as to obtain the prior distribution law of the collected images under this working condition. The image contains two parts: white speckles and black background, and the gray values of each pixel point gather around these two values respectively. The prior function is fitted with a distribution having bimodal characteristics. Generally, 10 images are collected for the fitting of the prior distribution law.

[0053] Further preferably, in step S4, use the prior distribution obtained in step S3 to perform super-resolution reconstruction on each group of images to obtain a high-resolution image. The magnification should be the multiple obtained after debugging in step S2, so that the number of speckles in the subset is appropriate and clear.

[0054] Further preferably, in step S4, the collection time interval of each group of image sequences used to obtain the high-resolution image is adjustable, and the number of images in the sequence can be adjusted according to the required quality of the reconstructed image and the acceptable reconstruction calculation duration.

[0055] Further preferably, in step S3, the fitting of the pixel gray value distribution is performed as follows:

[0056] The gray values of each pixel in the image of the collected specimen are statistically analyzed to obtain the proportion of the number of pixels with each gray value, and a frequency distribution curve is obtained. A function in the form of a double peak is used to fit it, and the function expression is as follows:

[0057]

[0058] In the formula, μ0 and μ1 represent the gray values at the two aggregation peaks, σ0 and σ1 represent the standard deviations of the two peaks, and H i,j represents the pixel value at the i-th row and j-th column in the predicted reconstructed high-resolution image H, and c is a normalization constant.

[0059] Further preferably, in step S4, the two cameras simultaneously collect a set of specimen image sequences. The acquisition time interval between adjacent images in the sequence is adjustable, and the number of images included in a set of sequences is adjustable, generally 7 images at an interval of 0.5 s. In this embodiment, a computer software analysis platform is used for SR reconstruction with a magnification of 2.0 times. A set of image sequences collected by the left and right cameras are respectively reconstructed into a high-resolution image.

[0060] The reconstruction algorithm adopts the maximum a posteriori probability (MAP) algorithm, and the prior function uses the double normal distribution function fitted in step three. The MAP estimator can be obtained by maximizing the posterior probability H MAP-DIC . The function expression is as follows:

[0061]

[0062] In the formula, L k represents the low-resolution image sequence before reconstruction, H represents the high-resolution image, and P DIC [H i,j represents the prior distribution function obtained in step three.

[0063] After the reconstruction based on the maximum a posteriori probability, a set of image sequences collected by the left and right cameras are respectively reconstructed into a high-resolution image, and the length and width resolutions both become 2.0 times the original.

[0064] Further preferably, in step S5, the high-resolution images corresponding to the left and right cameras are input into the DIC analysis software for calculation. After setting the required subset size and search step size and importing the previously calculated calibration data, 3D-DIC analysis is performed on the images collected by the two CCD cameras to obtain the three-dimensional morphology of the specimen 2 to be tested. This 3D-DIC analysis belongs to the existing technical means and will not be elaborated here.

[0065] The structure of the measurement system applicable to the above measurement method in the present invention is set as follows:

[0066] (1) Layout of two cameras

[0067] Two CCD cameras are symmetrically installed on the left and right sides of the gantry structure built by aluminum profiles, keeping the two CCD cameras on the same horizontal line. The distances from the lenses of the two cameras to the specimen along the optical axis are equal. An LCD projector is installed exactly in the middle of the two CCD cameras; as long as the optical axes of the two cameras are not parallel, the projected image by the projector is a speckle image;

[0068] After determining the distance from the camera to the specimen surface, on the premise of ensuring that the distances from the lenses of the two CCD cameras to the specimen along the optical axis are the same, adjust the distance between the two CCD cameras and the spatial solid angle (15° - 35°) so that the specimen appears completely and clearly in the fields of view of the two cameras;

[0069] (2) Focal length and angle of the projector

[0070] Determine the distance from the projector to the specimen surface, adjust the angle of the projector so that the projected pattern covers the specimen surface. Adjust the focal length of the projector so that a clear speckle pattern is formed in the region of interest on the surface of the specimen to be tested;

[0071] (3) Camera calibration

[0072] Reasonably select the position of the calibration plate so that the calibration plate appears clearly and completely in the fields of view of the two cameras at the same time. Change the attitude of the calibration plate, and the two cameras simultaneously take images of the calibration plate, and then calibrate the two cameras with the calibration algorithm;

[0073] (4) Setting of the speckle diameter

[0074] Use the computer software platform to set the diameter size of the generated speckle pattern, and then project the pattern onto the specimen to be tested.

[0075] The present invention will be further described below in conjunction with specific embodiments.

[0076] The specimen in this example is a 60mm × 60mm stepped block, machined by a CNC milling machine, and the dimensional parameters are as Figure 4 shown. Measure the morphology and step height of the stepped specimen to simulate the deformation measurement of an electronic packaging substrate and verify the accuracy of the morphology measurement.

[0077] The three-dimensional morphology measurement system in this example, as Figure 2 shown, includes: an optical experimental platform 1, a specimen to be tested 2, a CCD camera 3, an LCD projector 4, and a computer software analysis platform 5; among them, the CCD camera 3 includes a left camera L and a right camera R.

[0078] The test sample 2 is placed on the optical bench; the CCD camera 3 and the LCD projector 4 are installed on the gantry structure built by aluminum profiles for fixing the camera and the LCD projector; the computer software analysis platform 5 is connected to the CCD camera 3 and the LCD projector 4 for collecting images and generating projection speckle patterns; then the computer software analysis platform 5 is used for SR reconstruction of the images, and DIC analysis and calculation are performed using the reconstructed high-resolution images to obtain the surface topography information of the sample.

[0079] The specific implementation steps of a method for improving the accuracy of projection speckle digital image correlation deformation measurement according to the present invention are as follows:

[0080] Step 1: Installation of the camera and the projector, and calibration of the camera.

[0081] Step 1.1: Layout and installation of two cameras and a projector

[0082] Build the aluminum profile gantry structure on the optical experiment bench 1, and install the two CCD cameras 3 and the LCD projector 4 on the aluminum profiles. For the size of the stepped topography block sample 2 used in the test, adjust the distance between the camera and the sample surface to ensure that the two cameras are symmetrically distributed on both sides of the sample and the distances from the sample along the optical axis direction are the same. Adjust the camera spacing and spatial angle so that the sample is presented in the center of the fields of view of the two cameras and the sample images are complete and clear. Adjust the height and focal length of the projector so that the speckle pattern completely covers the area of interest of the sample to be tested, and an image with strong contrast and clear pattern is presented on the computer software platform 5.

[0083] Step 1.2: Calibration of the camera

[0084] Reasonably select the position of the calibration plate so that the calibration plate appears clearly and completely in the fields of view of the two CCD cameras 3 at the same time. Change the attitude of the calibration plate, and the computer software platform 5 collects and stores the images, and then calibrates the two cameras with the calibration algorithm.

[0085] The specific implementation method of Step 2 is as follows:

[0086] Step 2.1: Determine the speckle diameter

[0087] Use the computer software platform 5 to adjust the diameter of the generated speckle pattern. The size of the tested stepped block sample 2 is small and the step height is low, so a smaller speckle diameter is used to make the sample surface contain more gray-scale change information for analysis. Adopt a speckle pattern with a diameter of 3.0 pixels, as Figure 3 shown, and then project the pattern onto the sample to be tested.

[0088] Step 2.2: Determine the SR reconstruction magnification

[0089] The image of the test sample 2 covered with speckles is collected and stored, and then SR reconstruction is performed using the computer software platform 5. Adjust the magnification of the SR reconstruction so that in the reconstructed image, a square subset with a side length of 33 pixels contains 3 - 5 speckles. At this time, the SR magnification is 2.0 times.

[0090] The specific implementation method of step three is as follows:

[0091] Step 3.1: Count the number of pixels for each gray value

[0092] Collect 10 sample images and store them in the computer software platform 5, export the gray values of each pixel in these images, and count the distribution of the number of pixels for each gray value.

[0093] Step 3.2: Fit the prior function

[0094] After obtaining the bimodal distribution curve, use the double normal distribution function for fitting and obtain the expression of the distribution function.

[0095] The specific implementation method of step four is as follows:

[0096] Step 4.1: Dual - camera image acquisition and reconstruction

[0097] The two cameras simultaneously collect a sequence of 7 sample images, with a time interval of 0.5 s between each image acquisition. Then, use the computer software analysis platform 5 for SR reconstruction with a magnification of 2.0 times. Reconstruct a group of image sequences collected by the left and right cameras into a single high - resolution image respectively. The reconstruction algorithm uses the maximum a posteriori probability (MAP) algorithm, and the prior function uses the double normal distribution function fitted in step three.

[0098] Step 4.2: Digital image correlation analysis calculation

[0099] Input the high - resolution images corresponding to the left and right cameras into the DIC analysis software for calculation. Set the subset size to 33×33 pixels and the search step size to 15 pixels. After importing the previously calculated calibration data, perform 3D - DIC analysis on the images collected by the two CCD cameras to obtain the three - dimensional topography of the test sample 2, as Figure 7 shown.

[0100] Those skilled in the art can easily understand that the above - mentioned is only the preferred embodiment of the present invention, and it is not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting warping of an electronic packaging substrate combined with super-resolution reconstruction, characterized in that, The method includes the following steps: S1 Project a circular speckle pattern onto the surface of the test sample, collect multiple pictures of the projected pattern on the surface of the test sample, count the gray values of each pixel point in all the pictures, thereby obtaining the proportion P of the number of pixel points corresponding to each gray value to the total number of pixel points, draw a two-dimensional graph of the gray value and the proportion P corresponding to each gray value, fit it into a bimodal distribution pattern according to this two-dimensional graph and obtain the function of the proportion P, that is, the prior function; S2 Replace the test sample in step S1 with the sample to be measured, collect multiple photos of the projected pattern on the surface of the sample to be measured and obtain the gray values of all pixel points on each picture, use the prior function in step S1, and perform super-resolution reconstruction on the multiple collected photos according to a preset magnification factor, thereby obtaining a high-resolution image; S3 Calculate the three-dimensional topography of the surface of the sample to be measured by using the high-resolution image obtained in step S3, thereby obtaining the warping deformation information of the sample to be measured; In step S2, the prior function is as follows: where μ0 and μ1 are the gray values at the two clustering peaks respectively, σ0 and σ1 are the standard deviations of the two peaks respectively, H i,j is the pixel value at the i-th row and j-th column in the predicted reconstructed high-resolution image H, c is a normalization constant, P DIC [H i,j is the prior distribution function; In step S2, the reconstruction uses the maximum a posteriori probability algorithm, and the a posteriori probability function is as follows: Among them, L k is the low-resolution image sequence before reconstruction, H is the high-resolution image, and P DIC [H i,j is the prior distribution function, and H i,j is the pixel value at the i-th row and j-th column in the predicted reconstructed high-resolution image H; In steps S1 and S2, the photos of the test sample and the sample to be measured are both images collected from the left and right sides of the test sample or the sample to be measured respectively. In step S2, when reconstructing the high-resolution image, the images collected from the left and right sides are reconstructed separately, thereby obtaining the corresponding high-resolution images respectively.

2. The method for detecting the warping of an electronic packaging substrate combined with super-resolution reconstruction according to claim 1, wherein, In step S2, the preset magnification factor makes the reconstructed image contain 3 - 5 clear speckle spots in a square subset with a side length of 33 pixels.

3. A method for detecting the warping of an electronic packaging substrate combined with super-resolution reconstruction according to claim 1 or 2, characterized in that, In step S2, the high-resolution image refers to that the length and width resolution of the reconstructed image is between 1.5 times and 2.5 times that of the image before reconstruction.

4. A method for detecting warpage of an electronic packaging substrate combined with super-resolution reconstruction according to claim 1 or 2, characterized in that, In step S1, the speckle pattern is white spots on a black background, and the spot shape is circular.

5. A method for detecting warpage of an electronic packaging substrate combined with super-resolution reconstruction according to claim 1 or 2, characterized in that, In step S3, the 3D-DIC method is used to calculate the three-dimensional topography of the surface of the sample to be measured.

6. A measurement system for detecting the warping of an electronic packaging substrate combined with super-resolution reconstruction, characterized in that, The measurement system includes an image acquisition module, a prior function construction module, a super-resolution reconstruction module, and a three-dimensional topography construction module, where the image acquisition module is used to collect pictures of the circular speckle pattern projected on the test sample or the sample to be measured; the prior function construction module is used to construct a prior function according to the gray values of pixel points in the collected pictures; the prior function is as follows: where μ0 and μ1 are the gray values at the two clustering peaks respectively, σ0 and σ1 are the standard deviations of the two peaks respectively, and H i,j is the pixel value at the i-th row and j-th column in the predicted reconstructed high-resolution image H, c is a normalization constant, and P DIC [H i,j is the prior distribution function; the super-resolution reconstruction module is used to reconstruct the collected pictures into high-resolution images according to the prior function and the preset magnification factor by using the super-resolution reconstruction method; the reconstruction uses the maximum a posteriori probability algorithm, and the a posteriori probability function is as follows: Among them, L k is the low-resolution image sequence before reconstruction, H is the high-resolution image, and P DIC [H i,j is the prior distribution function, and H i,j is the pixel value at the i-th row and j-th column in the predicted reconstructed high-resolution image H; the three-dimensional topography construction module is used to obtain the three-dimensional topography of the surface of the sample to be measured by using the high-resolution image.

7. The measurement system according to claim 6, characterized in that, The measurement system further includes a control center. The image acquisition module includes a projector, a binocular camera, and an optical experimental platform. The projector and the binocular camera are arranged above the optical experimental platform. The two cameras in the binocular camera are respectively distributed on both sides of the projector. A test sample or a sample to be measured is placed on the optical experimental platform. The projector projects a circular speckle pattern onto the test sample or the sample to be measured.

Images