Automatic testing and sorting method for MTF (Modulation Transfer Function) value of chip

Abstract

The invention relates to the technical field of physical analysis, in particular to an automatic testing and sorting method for chip MTF values, which comprises the following steps: acquiring an MTF chart of a to-be-tested chip under each light source brightness in a plurality of cycle tests under different light source brightness, and determining the comprehensive reliability of each MTF chart, the light source brightness suitable factor of each MTF chart is determined by combining the curves in the MTF charts under adjacent different light source brightness in the same cycle test, and the superiority of each light source brightness is determined by combining the curves in the MTF charts under different cycle tests under the same light source brightness, so that the optimal MTF value of the chip to be tested is determined for sorting the chip to be tested. According to the method, the optimal light source brightness of the to-be-tested chip in the testing process is selected, so that the sorting accuracy of the to-be-tested chip is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of physical analysis, and particularly relates to an automatic test and sorting method for the MTF value of a chip. Background Art

[0002] MTF (Modulation Transfer Function) is the modulation transfer function, which is a term in the field of optical engineering and is used to evaluate the performance of a lens (optical system). It describes the ability of an optical system to transfer different spatial frequency components and reflects the degree to which the system preserves image details. The MTF value ranges from 0 to 1, and the higher the value, the stronger the ability of the optical system to preserve image details and the higher the imaging quality. Therefore, the chips can be sorted by obtaining the MTF value of the chip to be tested and setting a qualified threshold.

[0003] Existing problems: In preparing the automatic test environment for the MTF value of a chip, the selection of the light source brightness is one of the key factors affecting the test accuracy. Since the sensitivity (ISO value) of different chips is different, different brightnesses are required in the automatic test environment for the MTF value of different chips. An inappropriate light source brightness will result in inaccurate acquisition of the MTF value, thereby affecting the sorting of the chips. Summary of the Invention

[0004] The present invention provides an automatic test and sorting method for the MTF value of a chip to solve the existing problems.

[0005] The automatic test and sorting method for the MTF value of a chip of the present invention adopts the following technical solutions: An embodiment of the present invention provides an automatic test and sorting method for the MTF value of a chip, and the method includes the following steps: Obtain MTF charts for each light source brightness in several cyclic tests of the chip to be tested under different light source brightnesses; the MTF charts include curves of 30 lines in the tangential direction, curves of 30 lines in the radial direction, curves of 10 lines in the tangential direction, and curves of 10 lines in the radial direction; the horizontal axis of the MTF chart represents the distance from the monitoring point to the optical center of the lens, and the vertical axis represents the contrast; Determine the reliability of each curve according to the peaks and valleys on each curve in each MTF chart; Determine the comprehensive reliability of each MTF chart according to the reliability of each curve in each MTF chart and the difference between curves of the same size in different directions; Determine the light source brightness suitability factor of each MTF chart according to the comprehensive reliability of each MTF chart and the curves in MTF charts under adjacent different light source brightnesses in the same cyclic test; Determine the excellence of each light source brightness according to the light source brightness suitability factor of each MTF chart and the curves in the MTF chart in different cycle tests under the same light source brightness; Determine the optimal MTF value of the chip to be tested according to the excellence of each light source brightness; sort the chips to be tested according to the optimal MTF value of the chip to be tested.

[0006] Furthermore, the specific steps for determining the reliability of each curve are as follows: Denote any one curve in any one MTF chart as curve A; Divide curve A into several curve segments by the peak and valley in curve A; Calculate the difference between the maximum contrast and the minimum contrast in the th curve segment of curve A, and denote the ratio of the difference to the size of the horizontal axis range where the th curve segment of curve A is located as the first ratio; Calculate the product of the reciprocal of the median in the horizontal axis range where the th curve segment of curve A is located and the first ratio, and take it as the first product of the th curve segment of curve A; Determine the reliability of curve A according to the first product of each curve segment of curve A and all the contrasts in curve A.

[0007] Furthermore, the specific steps for determining the reliability of curve A are as follows: Calculate the product of the sum value of the first products of all the curve segments of curve A and the variance of all the contrasts in curve A, and denote it as the second product; Take the product of the inverse normalization value of the second product and the mean value of all the contrasts in curve A as the reliability of curve A.

[0008] Furthermore, the specific steps for determining the comprehensive reliability of each MTF chart are as follows: Denote the two curves of the 30 lines in the tangential and radial directions in any one MTF chart as curve A1 and curve A2 respectively; Calculate the normalized value of the absolute value of the difference between the ordinates corresponding to each same abscissa of curve A1 and curve A2, and take it as the difference value corresponding to each abscissa, and form a difference curve with all the difference values corresponding to the abscissas; On the difference curve, divide several curve segments with difference values greater than the preset first threshold, and denote them as abnormal curve segments; Denote the mean value of the reliabilities of curve A1 and curve A2 as the first mean value; Determine the comprehensive reliability of curve A1 and curve A2 according to the first mean value and each abnormal curve segment on the difference curve; In any one of the MTF charts, according to the method for obtaining the comprehensive reliability of the two curves of the 30-line in the tangential and radial directions, obtain the comprehensive reliability of the two curves of the 10-line in the tangential and radial directions, and take the average value of the comprehensive reliability of the two curves of the 30-line in the tangential and radial directions and the comprehensive reliability of the two curves of the 10-line in the tangential and radial directions as the comprehensive reliability of any one of the MTF charts.

[0009] Further, the specific steps for determining the comprehensive reliability of curve A1 and curve A2 are as follows: On the difference curve, calculate the product of the normalization value of the horizontal axis range size of the th abnormal curve segment and the maximum difference value in the th abnormal curve segment as the third product, calculate the product of the reciprocal of the median in the horizontal axis range of the th abnormal curve segment and the third product as the fourth product of the th abnormal curve segment, and take the product of the inverse normalization value of the sum of the fourth products of all abnormal curve segments and the first average value as the comprehensive reliability of curve A1 and curve A2.

[0010] Further, the specific steps for determining the light source brightness suitability factor of each MTF chart are as follows: In any one cycle test, sort all types of light source brightnesses from small to large, and sequentially obtain the MTF charts of the chip under test for each light source brightness according to the sorting to form an MTF chart sequence; In the MTF chart sequence, calculate the Pearson correlation coefficient between the curves of the 30-line in the tangential direction in the rd and the th MTF charts as the first correlation coefficient, calculate the Pearson correlation coefficient between the curves of the 30-line in the radial direction in the rd and the th MTF charts as the second correlation coefficient, calculate the normalization value of the sum of the first correlation coefficient and the second correlation coefficient as the first sum value, calculate the absolute value of the difference in the comprehensive reliability of the two curves of the 30-line in the tangential and radial directions in the rd and the th MTF charts as the first difference value, and take the ratio of the first sum value to the first difference value as the curve similarity of the 30-line in the tangential and radial directions in the rd and the th MTF charts ; According to the obtaining method, obtain the rd and the The curve similarity of the 10 lines in the tangential and radial directions in the Zhang MTF chart and the th and the curve similarity of the 10 lines in the tangential and radial directions in the Zhang MTF chart as well as the th and the curve similarity of the 30 lines in the tangential and radial directions in the Zhang MTF chart ; Take the , , and mean value as the second mean value, and multiply the second mean value by the comprehensive reliability of the th MTF chart in the MTF chart sequence as the light source brightness suitability factor of the th MTF chart in the MTF chart sequence.

[0011] Furthermore, the specific steps for determining the excellence of each light source brightness are as follows: In all cyclic tests, obtain all the MTF charts corresponding to the th light source brightness as the reference MTF charts; Determine the credibility of each data point on each curve in each reference MTF chart according to the trend change of each curve in each reference MTF chart; Denote the curves of the 30 lines in the tangential direction in any two reference MTF charts as curve A4 and curve A5 respectively; Use the DTW algorithm to match curve A4 and curve A5 to obtain several matching pairs and the DTW distance of each matching pair; Calculate the mean value of the credibility of the two data points in the th matching pair between curve A4 and curve A5 as the third mean value, and multiply the third mean value by the DTW distance of the th matching pair between curve A4 and curve A5 as the fifth product of the th matching pair between curve A4 and curve A5. Take the sum value of the fifth products of all the matching pairs between curve A4 and curve A5 as the initial inconsistency between curve A4 and curve A5; For any two reference MTF charts, obtain the initial inconsistency of the curve of the 30 lines in the radial direction in the same way as the acquisition method of the initial inconsistency of the curve of the 30 lines in the tangential direction , the initial inconsistency of the curve of the 10 lines in the tangential direction and the initial inconsistency of the curve of the 10 lines in the radial direction , and take the , , and The inverse proportional normalization value of the sum value is used as the overall shape consistency of the curves of any two reference MTF charts; The mean value of the overall shape consistency of the curves of all any two reference MTF charts is denoted as the fourth mean value, the mean value of the light source brightness suitability factors corresponding to all reference MTF charts is denoted as the fifth mean value, and the product of the fourth mean value and the fifth mean value is used as the goodness of the

[0012] Furthermore, the determination of the credibility of each data point on each curve in each reference MTF chart includes the following specific steps: Denote any curve in any reference MTF chart as curve A3; Divide curve A3 into several curve segments with the peak and trough in curve A3, and denote the curve segment with the starting point as the trough or the ending point as the peak as the upward trend curve segment; Calculate the difference between the maximum contrast and the minimum contrast in the th upward trend curve segment of curve A3, and denote the ratio of the difference to the size of the horizontal axis range where the th upward trend curve segment of curve A3 is located as the second ratio. Calculate the product of the reciprocal of the median value within the horizontal axis range where the th upward trend curve segment of curve A3 is located and the second ratio, and use it as the data unbelievability of the th upward trend curve segment of curve A3 ; Use as the credibility of each data point in the th upward trend curve segment of curve A3; where is a linear normalization function; Set the credibility of each data point on curve A3 that is not in the upward trend curve segment to a preset second threshold.

[0013] Furthermore, the determination of the optimal MTF value of the chip under test includes the following specific steps: Among the goodness of all light source brightnesses, use the light source brightness corresponding to the maximum goodness as the optimal light source brightness of the chip under test; Obtain the MTF values at the preset specific spatial frequencies in all MTF charts of the chip under test under the optimal light source brightness, and use the mean value of the MTF values at the preset specific spatial frequencies in all MTF charts as the optimal MTF value of the chip under test.

[0014] Further, the sorting of the chip to be tested includes the following specific steps: When the optimal MTF value is greater than the preset qualified threshold, it is determined that the chip to be tested meets the standard; When the optimal MTF value is less than or equal to the preset qualified threshold, it is determined that the chip to be tested does not meet the standard.

[0015] The beneficial effects of the technical solution of the present invention are: In the embodiment of the present invention, MTF charts under different light source brightnesses in several cyclic tests of the chip to be tested are obtained, and the comprehensive reliability of each MTF chart is determined. Thus, by analyzing the changes in the curves in the MTF chart and the differences in the curves at the same spatial frequency, it is judged whether the MTF chart is reliable, which is used as the basic parameter for selecting the optimal light source brightness. Combining the curves in the MTF charts under adjacent different light source brightnesses in the same cyclic test, the light source brightness suitability factor of each MTF chart is determined. Thus, by analyzing the similarity of the curves in the MTF charts under different light source brightnesses, the accuracy of selecting the optimal light source brightness is ensured. Then, combining the curves in the MTF charts in different cyclic tests under the same light source brightness, the excellence of each light source brightness is determined. Thus, by analyzing the shape consistency of the curves in the MTF charts in different cyclic tests under the same light source brightness, the accuracy of selecting the optimal light source brightness is further ensured. Finally, according to the optimal light source brightness, the optimal MTF value of the chip to be tested is determined, which is used to sort the chip to be tested. So far, the present invention ensures the accuracy of sorting the chip to be tested by selecting the optimal light source brightness during the test process. Description of the Drawings

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0017] Figure 1 It is a step flow chart of an automatic test and sorting method for the MTF value of a chip of the present invention; Figure 2 It is a schematic diagram of the moving distance between the test card and the optical center of the lens; Figure 3 It is an MTF chart. Detailed Embodiments

[0018] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following specifically describes, in conjunction with the accompanying drawings and preferred embodiments, a method for automatically testing and sorting the MTF value of a chip according to the present invention, including its specific implementation manner, structure, features, and effects. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0020] The following specifically describes, in conjunction with the accompanying drawings, the specific solution of a method for automatically testing and sorting the MTF value of a chip provided by the present invention.

[0021] Please refer to Figure 1 , which shows the step flow chart of a method for automatically testing and sorting the MTF value of a chip provided by an embodiment of the present invention. The method includes the following steps: Step S001: Obtain MTF charts for each light source brightness in several cyclic tests of the chip to be tested under different light source brightnesses; the MTF charts include curves of 30 lines in the tangential direction, curves of 30 lines in the radial direction, curves of 10 lines in the tangential direction, and curves of 10 lines in the radial direction; the horizontal axis of the MTF chart represents the distance from the monitoring point to the optical center of the lens, and the vertical axis represents the contrast.

[0022] The purpose of this embodiment is to obtain the optimal MTF value from the MTF charts of the chip under different light source brightnesses to ensure the accuracy of chip sorting.

[0023] Set up a stable test environment, including a light source, the chip to be tested, an imaging device, etc. Under the condition that other conditions remain unchanged, for the same chip to be tested, preset several different light source brightnesses, and sequentially obtain the MTF charts for each light source brightness from small to large as a cyclic test, and obtain the MTF charts of the chip to be tested for each light source brightness in several cyclic tests.

[0024] It should be noted that: in this embodiment, the range of the light source brightness is 50 to 300 lumens (lm), and a test is performed every 10 lumens, and the number of cycles of the cyclic test is 10. This is used as an example for description.

[0025] The specific process of obtaining the MTF chart is as follows: In this embodiment, test cards with 30 lines and 10 lines are used. The contrast of the imaging is obtained every 5 millimeters (mm) from the center of the screen to the edge, that is, in the direction from the center of the screen to the edge, the center of the screen, 5 mm, 10 mm, 15 mm, and 20 mm away from the center of the screen are used as monitoring points respectively, and the contrast of the imaging at each monitoring point is obtained. The schematic diagram of the moving distance between the test card and the optical center of the lens is as shown in Figure 2 shown. The MTF chart is as shown in Figure 3 shown.

[0026] It should be noted that: Figure 2 In [the figure], the thin lines of 30 lines are used to test the resolution (resolving power), and the thick lines of 10 lines are used to test the contrast. V2 is the tangential line (the line perpendicular to the direction from the center of the screen to the edge), which is represented as a dotted line (U1 and U3) on the MTF chart. V1 is the radial line (the line parallel to the direction from the center of the screen to the edge), which is represented as a solid line (U2 and U4) on the MTF chart. Among them, the radial and tangential directions can help understand the imaging performance of the optical system in different directions. Therefore, each MTF chart contains the curve of 30 lines in the tangential direction, the curve of 30 lines in the radial direction, the curve of 10 lines in the tangential direction, and the curve of 10 lines in the radial direction, as shown in Figure 3 shown, where Figure 3 the four curves in [the figure] from top to bottom are the curve of 10 lines in the tangential direction represented by the dotted line U1, the curve of 10 lines in the radial direction represented by the solid line U2, the curve of 30 lines in the tangential direction represented by the dotted line U3, and the curve of 30 lines in the radial direction represented by the solid line U4. Figure 3 In [the figure], the horizontal axis of the MTF chart represents the distance between the monitoring point and the optical center of the lens, with the unit of millimeter (mm), and the vertical axis represents the contrast, with the unit of percentage (%). Among them, Figure 2 the center of the screen in [the figure] represents the optical center of the lens in Figure 3 . The MTF chart is used to show the contrast transfer ability of the optical system at different distances, which helps to evaluate the spatial resolution and imaging quality of the system.

[0027] Step S002: Determine the reliability of each curve according to the peaks and valleys on each curve in each MTF chart.

[0028] In the MTF chart, the higher the contrast of the curve, the more accurately the imaging chip can capture the details of the pattern. And the smoother the curve, the more uniform the change of the imaging performance of the optical system during the process of moving away from the optical center of the lens, without obvious performance fluctuations or mutations. This indicates that the contrast transfer ability of the system when processing different detail sizes is relatively stable, which is conducive to providing consistent image quality. The smooth curve may also mean that the noise and errors introduced during the test are less, making the data more reliable.

[0029] Preferably, in an embodiment of the present invention, the method for obtaining the reliability of each curve in each MTF chart includes: Denote any one curve in any one MTF chart as curve A.

[0030] Use the peak-valley detection algorithm to obtain the peaks and valleys in curve A, and divide curve A into several curve segments based on the peaks and valleys in curve A. Among them, the peak-valley detection algorithm is a well-known technology, and the specific method will not be introduced here.

[0031] Calculate the difference between the maximum contrast and the minimum contrast in the th curve segment of curve A, and denote the ratio of this difference to the size of the horizontal axis range where the th curve segment of curve A is located as the first ratio. Calculate the product of the reciprocal of the median value within the horizontal axis range where the th curve segment of curve A is located and the first ratio, and take it as the first product of the th curve segment of curve A. Calculate the product of the sum value of the first products of all curve segments of curve A and the variance of all contrasts in curve A, and denote it as the second product. Take the product of the inverse proportional normalization value of the second product and the mean value of all contrasts in curve A as the reliability of curve A.

[0032] It should be noted that the calculation formula for the reliability of curve A is: In the formula, is the reliability of curve A, is the mean value of all contrasts in curve A. The larger is, the more accurately the details of the pattern can be captured. is the variance of all contrasts in curve A. The smaller is, the greater the overall smoothness of curve A, and the more reliable curve A is. is the number of curve segments into which curve A is divided. is the difference between the maximum contrast and the minimum contrast in the th curve segment of curve A. represents the size of the horizontal axis range where the th curve segment of curve A is located. is the first ratio. The larger is, the greater the contrast change appears under a smaller distance change, that is, a large contrast mutation occurs locally where the th curve segment of curve A is located. is the median value within the horizontal axis range where the The smaller it is, the closer it is to the optical center of the lens. The center of the lens is usually the area with the highest imaging quality. The imaging quality of the central area directly reflects the basic performance of the lens. That is, the imaging quality analysis closer to the lens center is more important. Therefore, is used as the weight, is the first product of the th curve segment of curve A. The larger it is, the smaller the local smoothness of curve A indicates. is the second product. is the exponential function with the natural constant as the base. In this embodiment, is used to present the inverse proportional relationship and normalization processing of the second product. The implementer can set the inverse proportional function and normalization function according to the actual situation. Therefore, represents the comprehensive smoothness of curve A. The larger its value, the smoother it is. Therefore, the larger it is, the more reliable curve A is.

[0033] Step S003: Determine the comprehensive reliability of each MTF chart according to the reliability of each curve in each MTF chart and the differences between curves of the same-sized lines in different directions.

[0034] Since 30 lines usually represent higher spatial frequencies, that is, the details of the image, while 10 lines represent lower spatial frequencies, that is, the larger features of the image. At the same spatial frequency, when the curves in different directions are more consistent, it means that the lens shows similar capabilities in processing details in different directions, which means that the image maintains good contrast and detail retention capabilities throughout the field of view.

[0035] Preferably, in an embodiment of the present invention, the method for obtaining the comprehensive reliability of each MTF chart includes: Denote the two curves of 30 lines in the tangential and radial directions in any MTF chart as curve A1 and curve A2 respectively.

[0036] Calculate the normalized value of the absolute value of the difference between the ordinates corresponding to the same abscissa of curve A1 and curve A2 as the difference value corresponding to each abscissa, and form a difference curve with the difference values corresponding to all abscissas.

[0037] Among them, for the normalized value of the absolute value of the difference, in this embodiment, a linear normalization function is used to normalize the absolute value of the difference to between 0 and 1.

[0038] Preset the first threshold to 0.2 and describe it by taking this as an example.

[0039] On the difference curve, divide several curve segments with difference values greater than the preset first threshold and denote them as abnormal curve segments.

[0040] Denote the mean of the reliability of curve A1 and curve A2 as the first mean.

[0041] On the difference curve, calculate the product of the normalized value of the horizontal axis range where the th abnormal curve segment is located and the maximum difference value in the th abnormal curve segment as the third product. Calculate the product of the reciprocal of the median within the horizontal axis range where the th abnormal curve segment is located and the third product as the fourth product of the th abnormal curve segment. Take the product of the inverse proportional normalized value of the sum of the fourth products of all abnormal curve segments and the first mean as the comprehensive reliability of curve A1 and curve A2.

[0042] It should be noted that: The calculation formula for the comprehensive reliability of curve A1 and curve A2 is: In the formula, is the comprehensive reliability of curve A1 and curve A2, and are the reliabilities of curve A1 and curve A2 respectively, is the first mean, is the number of abnormal curve segments in the difference curve, is the th median within the horizontal axis range where the abnormal curve segment is located, is the th horizontal axis range size where the abnormal curve segment is located, is the horizontal axis range size of the difference curve. In this embodiment, use as the normalized value, is the th maximum difference value in the abnormal curve segment, is the third product. The larger the third product, the more inconsistent the imaging performance of the 30-line in the radial and tangential directions within a larger distance range, indicating that the data is less reliable. Using as the weight, The smaller it is, the closer it is to the optical center of the lens. Usually, the center of the lens is the area with the highest imaging quality. The imaging quality of the central area directly reflects the basic performance of the lens, that is, the imaging quality analysis closer to the center of the lens is more important. is the th fourth product of the abnormal curve segment to obtain the adjustment coefficient of the reliability, so as to obtain the comprehensive reliability of the 30-line curve in the radial and tangential directions, is the exponential function with the natural constant as the base. In this embodiment, use to present For the inverse proportional relationship and normalization processing, the implementer can set the inverse proportional function and normalization function according to the actual situation.

[0043] In any MTF chart, according to the method of obtaining the comprehensive reliability of the two curves of the 30 - line under tangential and radial directions, obtain the comprehensive reliability of the two curves of the 10 - line under tangential and radial directions, and take the mean value of the comprehensive reliability of the two curves of the 30 - line under tangential and radial directions and the comprehensive reliability of the two curves of the 10 - line under tangential and radial directions as the comprehensive reliability of any MTF chart.

[0044] Step S004: Determine the light source brightness suitability factor for each MTF chart according to the comprehensive reliability of each MTF chart and the curves in the MTF chart under different light source brightnesses adjacent in the same cycle test.

[0045] Preferably, in an embodiment of the present invention, the method for obtaining the light source brightness suitability factor for each MTF chart includes: In any cycle test, sort all kinds of light source brightnesses from small to large, and obtain the MTF charts of the chip under test for each light source brightness in sequence according to the sorting to form an MTF chart sequence.

[0046] Since the dynamic range of the imaging system is limited, it defines the ratio between the brightest and darkest signals that the system can handle. Within the normal brightness range, the dynamic range of the system can fully cope with the brightness change, so the change of the MTF chart is small. While in the case of too low or too high brightness, the system may be close to the limit of its dynamic range, and at this time, the change of brightness may cause a significant change in the MTF chart. Therefore, in the MTF chart sequence, when a certain MTF chart is more similar to its adjacent MTF chart, it indicates that the light source brightness corresponding to this certain MTF chart is more suitable for the chip under test.

[0047] In the MTF chart sequence, calculate the Pearson correlation coefficient between the curves of the 30 - line under the tangential direction of the th and the th MTF charts as the first correlation coefficient, calculate the Pearson correlation coefficient between the curves of the 30 - line under the radial direction of the th and the th MTF charts as the second correlation coefficient, calculate the normalized value of the sum of the first correlation coefficient and the second correlation coefficient as the first sum value, calculate the absolute value of the difference between the comprehensive reliabilities of the two curves of the 30 - line under the tangential and radial directions of the th and the th MTF charts as the first difference, and take the ratio of the first sum value to the first difference as the th and the Curve similarity of the 30 - line in the tangential and radial directions in the MTF chart 。

[0048] It should be noted that: Among the MTF chart sequences, the th and the th MTF charts, the curve similarity of the 30 - line in the tangential and radial directions The calculation formula is: In the formula, is the Pearson correlation coefficient between the curves of the 30 - line in the tangential direction in the th and the th MTF charts in the MTF chart sequence, and it is also the first correlation coefficient. is the Pearson correlation coefficient between the curves of the 30 - line in the radial direction in the th and the th MTF charts in the MTF chart sequence, and it is also the second correlation coefficient. Among them, the Pearson correlation coefficient is a well - known technology. The Pearson correlation coefficient is between - 1 and 1. The closer it is to 1, the more consistent the change trends of the two curves are, indicating that the performance of the optical system is relatively stable under these two lighting conditions. Therefore, the light source brightness corresponding to the th MTF chart is more suitable for the chip to be measured. is the linear normalization function, and are respectively the comprehensive reliabilities of the two curves of the 30 - line in the tangential and radial directions in the th and the th MTF charts. is the absolute - value function, The smaller it is, the more similar the comprehensive reliabilities of the curves of the 30 - line in the radial and tangential directions in the two MTF charts are. Therefore, The larger it is, the more similar the curves of the 30 - line in the radial and tangential directions between the two MTF charts are.

[0049] In the MTF chart sequence, in the above - mentioned manner, obtain the curve similarity of the 10 - line in the tangential and radial directions in the th and the th MTF charts, the curve similarity of the 10 - line in the tangential and radial directions in the th and the th MTF charts, and the curve similarity of the 30 - line in the tangential and radial directions in the th and the th MTF charts.

[0050] It should be noted that: The first and the last MTF charts in the MTF chart sequence are not analyzed. They are in the lowest and highest test light source brightness, which are extreme states and not used for the selection of appropriate light source brightness.

[0051] Take the mean values of , , and as the second mean value. Multiply the second mean value by the comprehensive reliability of the th MTF chart in the MTF chart sequence, and use the product as the light source brightness suitability factor of the th MTF chart in the MTF chart sequence.

[0052] It should be noted that: The greater the comprehensive reliability, the more reliable the th MTF chart. And the greater the second mean value, the more similar the th MTF chart is to its adjacent MTF images. Therefore, the greater the light source brightness suitability factor, the more suitable the light source brightness corresponding to the th MTF chart is for the chip under test.

[0053] Step S005: Determine the goodness of each light source brightness according to the light source brightness suitability factor of each MTF chart and the curves in the MTF charts during different cycle tests under the same light source brightness.

[0054] Since the brightness of the light source is directly affected by the supply voltage and current. Any slight power fluctuation may cause the fluctuation of the light source brightness, and the dust in the environment will affect the propagation and distribution of light, and the change of dust will also cause the brightness fluctuation.

[0055] Therefore, under a certain light source brightness, the credibility of a single test is relatively low. So in this embodiment, the optimal light source brightness is determined through the analysis of the MTF charts under multiple tests with the same light source brightness.

[0056] Preferably, in an embodiment of the present invention, the method for obtaining the goodness of each light source brightness includes: In all cycle tests, obtain all the MTF charts corresponding to the th light source brightness (i.e., the th MTF chart in each cycle test) as the reference MTF charts, and obtain multiple reference MTF charts.

[0057] Since in the MTF curve graph, theoretically the curve should show a downward trend as it moves away from the center of the lens, and as the distance from the center increases, the speed of the curve decline will accelerate because of the reduction of contrast caused by aberration and diffraction effects in the edge region.

[0058] Therefore, in the MTF chart, if there is a curve segment in an upward trend on a curve, the upward-trending curve segment may be caused by measurement errors, data processing errors, or other unexpected factors. The data in the upward segment is usually considered untrustworthy because it does not conform to the normal behavior of the optical system.

[0059] Take any curve in any reference MTF chart and denote it as curve A3. Obtain the curve segments of curve A3 divided according to peaks and valleys. Denote the curve segments with the starting point at a valley or the ending point at a peak as upward-trending curve segments.

[0060] Calculate the difference between the maximum contrast and the minimum contrast in the th upward-trending curve segment of curve A3. Denote the ratio of this difference to the size of the horizontal axis range where the th upward-trending curve segment of curve A3 is located as the second ratio. Calculate the product of the reciprocal of the median within the horizontal axis range where the th upward-trending curve segment of curve A3 is located and the second ratio, and take it as the untrustworthiness of the data in the th upward-trending curve segment of curve A3. .

[0061] It should be noted that: the smaller the median within the horizontal axis range where the th upward-trending curve segment of curve A3 is located, the closer the th upward-trending curve segment is to the lens center, and the more untrustworthy the data in the th upward-trending curve segment is. And the larger the second ratio, the greater the increase in contrast appears under a smaller distance change, that is, the more untrustworthy the data in the th upward-trending curve segment is. Therefore, the larger is, the more untrustworthy the data in the th upward-trending curve segment is.

[0062] Take as the credibility of each data point in the th upward-trending curve segment of curve A3.

[0063] Among them, is a linear normalization function used to normalize to between 0 and 1.

[0064] Obtain the credibility of each data point in each upward-trending curve segment of curve A3 in the above manner.

[0065] Preset the second threshold as 1 and use this as an example for description.

[0066] Set the credibility of each data point on curve A3 that is not in the upward trend curve segment to a preset second threshold, thereby obtaining the credibility of each data point on curve A3.

[0067] Since the overall shape of the curve in the MTF chart represents the imaging performance of the optical system.

[0068] Denote the curves of the 30 lines in the tangential direction in any two reference MTF charts as curve A4 and curve A5 respectively.

[0069] Use the DTW algorithm to match curve A4 and curve A5, obtaining a number of matching pairs and the DTW distance of each matching pair.

[0070] It should be noted that: The DTW algorithm is a well-known technology. The smaller the DTW distance, the more similar the two curves are. Each matching pair includes a data point in curve A4 and a data point in curve A5. The DTW distance of each matching pair is represented by the Euclidean distance between the two data points in each matching pair.

[0071] Calculate the mean value of the credibilities of the two data points in the th matching pair between curve A4 and curve A5 as the third mean value, and take the product of the third mean value and the DTW distance of the th matching pair between curve A4 and curve A5 as the fifth product of the th matching pair between curve A4 and curve A5. Take the sum value of the fifth products of all the matching pairs between curve A4 and curve A5 as the initial inconsistency between curve A4 and curve A5.

[0072] For any two reference MTF charts, obtain the initial inconsistency of the curve of the 30 lines in the radial direction in the same way as the acquisition method of the initial inconsistency of the curve of the 30 lines in the tangential direction , the initial inconsistency of the curve of the 10 lines in the tangential direction and the initial inconsistency of the curve of the 10 lines in the radial direction . Take the , , and sum value 's inverse proportional normalization value as the overall shape consistency of the curves of the two reference MTF charts.

[0073] It should be noted that: 's inverse proportional normalization value is represented by . Among them, the larger the third mean value, the The DTW distance of a matching pair is more reliable. That is, when analyzing the overall shape consistency of curves, focus on the data that conforms to the expected downward trend to obtain a reliable DTW distance. The larger the DTW distance, the more inconsistent the shapes of the two curves. Therefore, use The inverse proportional normalization value of is used as the overall shape consistency of the curves.

[0074] Therefore, in all cyclic tests, in each cyclic test, the th MTF chart (reference MTF chart), that is, the corresponding The specific process of obtaining the excellence of the light source brightness of the th kind is as follows: Denote the mean value of the overall shape consistency of the curves of all arbitrarily two reference MTF charts as the fourth mean value, denote the mean value of the light source brightness suitability factors corresponding to all reference MTF charts as the fifth mean value, and take the product of the fourth mean value and the fifth mean value as the excellence of the light source brightness of the th kind.

[0075] It should be noted that: the larger the fourth mean value, the more consistent the overall shapes of the curves of the MTF charts of all sub-tests under the light source brightness of the th kind, that is, the th kind of light source brightness has better stability in repeated tests. The larger the fifth mean value, the more suitable the th kind of light source brightness is for the chip to be tested. Therefore, the greater the excellence, the more suitable the light source brightness is for the chip to be tested.

[0076] Step S006: Determine the optimal MTF value of the chip to be tested according to the excellence of each kind of light source brightness; sort the chips to be tested according to the optimal MTF value of the chip to be tested.

[0077] Preferably, in an embodiment of the present invention, the obtaining method for sorting the chips to be tested includes: Obtain the excellence of each kind of light source brightness in the above manner, and use the light source brightness corresponding to the maximum excellence as the optimal light source brightness of the chip to be tested in this embodiment.

[0078] Obtain the MTF charts of the chip to be tested in multiple cyclic tests under the optimal light source brightness, obtain the MTF value of the chip to be tested at a preset specific spatial frequency from each MTF chart, and take the mean value of the MTF values at the preset specific spatial frequency in all MTF charts of the chip to be tested under the optimal light source brightness as the optimal MTF value of the chip to be tested.

[0079]

[0079] It should be noted that in this embodiment, the average value of the contrast of the two curves of 30 lines in the radial and tangential directions at a distance of 5 mm from the optical center of the lens in each MTF chart is used as the MTF value of the chip under test at a preset specific spatial frequency in each MTF chart. That is, the preset specific spatial frequency is 30 lines, and it includes two directions, radial and tangential. The selected horizontal axis coordinate is 5 mm from the optical center of the lens. This is used as an example for description.

[0080] Set the pass threshold of the MTF value according to the product quality requirements of the chip under test. In this embodiment, the preset pass threshold is 0.8. This is used as an example for description.

[0081] Compare the optimal MTF value of the chip under test with the preset pass threshold to determine whether it meets the standard. When the optimal MTF value is greater than the preset pass threshold, it is determined that the chip under test meets the standard. When the optimal MTF value is less than or equal to the preset pass threshold, it is determined that the chip under test does not meet the standard.

[0082] Classify the qualified chips into the qualified category and the unqualified chips into the unqualified category.

[0083] It should be noted that in this embodiment, when the denominator in the formula is 0, the denominator is set to 1. This is used as an example for description to ensure that the formula holds.

[0084] So far, the present invention is completed.

[0085] In summary, in the embodiment of the present invention, several MTF charts of the chip under test under different light source brightnesses in a number of cyclic tests are obtained, the comprehensive reliability of each MTF chart is determined, the light source brightness suitability factor of each MTF chart is determined by combining the curves in the MTF charts under adjacent different light source brightnesses in the same cyclic test, and then the excellence of each light source brightness is determined by combining the curves in the MTF charts in different cyclic tests under the same light source brightness, so as to determine the optimal MTF value of the chip under test for sorting the chip under test. The present invention ensures the accuracy of sorting the chip under test by selecting the optimal light source brightness during the test process.

[0086] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for automatically testing and sorting chip MTF values, characterized in that: The method comprises the following steps: Obtaining an MTF chart of the chip under test at each light source brightness in several cycle tests at different light source brightnesses; the MTF chart includes a curve of 30 lines under the tangential direction, a curve of 30 lines under the radial direction, a curve of 10 lines under the tangential direction, and a curve of 10 lines under the radial direction; the horizontal axis of the MTF chart represents the distance between the monitoring point and the optical center of the lens, and the vertical axis represents the contrast; Determine the reliability of each curve based on the peaks and valleys on each curve in each MTF chart; Determine the overall reliability of each MTF chart based on the reliability of each curve in each MTF chart and the differences in curves of lines of the same size in different directions; Determine the light source brightness suitability factor of each MTF chart based on the comprehensive reliability of each MTF chart and the curves in the MTF charts at adjacent different light source brightnesses in the same cycle test; Determine the quality of each light source brightness based on the light source brightness suitability factor of each MTF chart and the curves in the MTF chart in different cycle tests under the same light source brightness; According to the excellence of the brightness of each light source, the optimal MTF value of the chip to be tested is determined; according to the optimal MTF value of the chip to be tested, the chips to be tested are sorted.

2. According to claim 1, a method for automatically testing and sorting chip MTF values, characterized in that: The specific steps of determining the reliability of each curve are as follows: Let any curve in any MTF chart be recorded as curve A; Divide the curve A into several curve segments based on the peaks and valleys in the curve A; Calculate the curve A The difference between the maximum contrast and the minimum contrast in the first curve segment is compared with the first curve segment of curve A. The ratio of the horizontal axis range where the curve segments are located is recorded as the first ratio; Calculate the curve A The product of the reciprocal of the median value within the horizontal axis range of the first curve segment and the first ratio is taken as the first value of curve A. The first product of curve segments; Based on the first product of each curve segment of curve A and all contrasts in curve A, the reliability of curve A is determined.

3. According to claim 2, a method for automatically testing and sorting chip MTF values ​​is characterized in that: The reliability of the curve A is determined, and the specific steps include the following: Calculate the product of the sum of the first products of all curve segments of curve A and the variance of all contrasts in curve A, and record it as the second product; The product of the inverse normalized value of the second product and the mean of all contrasts in curve A is taken as the reliability of curve A.

4. According to claim 1, a method for automatically testing and sorting chip MTF values, characterized in that: The specific steps of determining the comprehensive reliability of each MTF chart are as follows: The two curves of the 30-degree line in the tangential and radial directions in any MTF chart are recorded as curve A1 and curve A2 respectively; Calculate the normalized value of the absolute value of the difference between the curve A1 and the curve A2 at each ordinate corresponding to the same abscissa as the difference value corresponding to each abscissa, and construct a difference curve with the difference values ​​corresponding to all abscissas; On the difference curve, a number of curve segments whose difference values ​​are greater than a preset first threshold are divided and recorded as abnormal curve segments; The mean of the reliability of curve A1 and curve A2 is recorded as the first mean; Determine the comprehensive reliability of curve A1 and curve A2 according to the first mean value and each abnormal curve segment on the difference curve; In any of the MTF charts, the comprehensive reliability of the two curves of 30 lines in the tangential and radial directions is obtained according to the method for obtaining the comprehensive reliability of the two curves of 10 lines in the tangential and radial directions, and the average of the comprehensive reliability of the two curves of 30 lines in the tangential and radial directions and the comprehensive reliability of the two curves of 10 lines in the tangential and radial directions is taken as the comprehensive reliability of any of the MTF charts.

5. According to claim 4, a method for automatically testing and sorting chip MTF values ​​is characterized in that: The specific steps of determining the comprehensive reliability of curve A1 and curve A2 are as follows: On the difference curve, calculate The normalized value of the horizontal axis range of the abnormal curve segment is The product of the maximum difference values ​​in the abnormal curve segments is used as the third product to calculate the The product of the reciprocal of the median value in the horizontal axis range where the abnormal curve segment is located and the third product is taken as the The fourth product of the abnormal curve segments is obtained by multiplying the inversely proportional normalized value of the sum of the fourth products of all the abnormal curve segments and the first mean value as the comprehensive reliability of curve A1 and curve A2.

6. According to claim 4, a method for automatically testing and sorting chip MTF values ​​is characterized in that: The specific steps of determining the light source brightness suitability factor of each MTF chart are as follows: In any cycle test, the brightness of all light sources is sorted from small to large, and the MTF chart of the chip to be tested under each light source brightness is obtained in sequence according to the sorting to form an MTF chart sequence; In the MTF chart sequence, calculate the Zhang Yudi The Pearson correlation coefficient between the curves of the 30 lines in the MTF chart is used as the first correlation coefficient. Zhang Yudi The Pearson correlation coefficient between the 30-line curves under the radial direction in the MTF chart is used as the second correlation coefficient. The normalized value of the sum of the first correlation coefficient and the second correlation coefficient is calculated as the first sum value. Zhang Yudi The absolute value of the difference between the comprehensive reliability of the two curves of the 30-line in the tangential and radial directions in the MTF chart is taken as the first difference, and the ratio of the first sum to the first difference is taken as the second Zhang Yudi Similarity of the curves of 30 lines in the tangential and radial directions in the MTF chart ; according to How to obtain the Zhang Yudi Similarity of the curves of the 10 lines in the tangential and radial directions in the MTF chart , No. Zhang Yudi Similarity of the curves of the 10 lines in the tangential and radial directions in the MTF chart and Zhang Yudi Similarity of the curves of 30 lines in the tangential and radial directions in the MTF chart ; Will , , as well as The second mean is taken as the mean of The product of the comprehensive reliability of the MTF charts is used as the first The light source brightness suitability factor for the MTF chart.

7. The automatic testing and sorting method of chip MTF value according to claim 1, characterized in that: The step of determining the quality of the brightness of each light source includes the following specific steps: In all loop tests, get the All MTF charts corresponding to the brightness of the light source are used as reference MTF charts; According to the trend change of each curve in each reference MTF chart, the credibility of each data point on each curve in each reference MTF chart is determined; The curves of the 30-degree line in the tangent direction in any two reference MTF charts are recorded as curve A4 and curve A5 respectively; Use the DTW algorithm to match curve A4 and curve A5 to obtain several matching pairs and the DTW distance of each matching pair; Calculate the first The third mean is compared with the first mean between curve A4 and curve A5. The product of the DTW distances of the matching pairs is used as the first The fifth product of each matching pair is calculated, and the sum of the fifth products of all matching pairs between curve A4 and curve A5 is taken as the initial inconsistency between curve A4 and curve A5; For any two reference MTF charts, the initial inconsistency of the curves at 30 degrees below the tangent line is The acquisition method obtains the initial inconsistency of the curve of the 30 lines under the radial direction , the initial inconsistency of the curve of the 10 lines under the tangent and the initial inconsistency of the curve of the 10 lines under the radial ,Will , , as well as The inversely proportional normalized value of the sum of the values ​​is used as the overall shape consistency of the curves of any two reference MTF charts; The mean of the overall shape consistency of the curves of all two reference MTF charts is recorded as the fourth mean, the mean of the light source brightness suitability factors corresponding to all reference MTF charts is recorded as the fifth mean, and the product of the fourth mean and the fifth mean is recorded as the fifth mean. The excellence of the brightness of the light source.

8. The automatic testing and sorting method of chip MTF value according to claim 7, characterized in that: The specific steps of determining the credibility of each data point on each curve in each reference MTF chart are as follows: Denote any curve in any reference MTF chart as curve A3; The curve A3 is divided into several curve segments according to the peak and valley in the curve A3, and the curve segment whose starting point is the valley or whose end point is the peak is recorded as the rising trend curve segment; Calculate the curve A3 The difference between the maximum contrast and the minimum contrast in the first rising trend curve segment is compared with the first The ratio of the horizontal axis range of the first rising trend curve segment is recorded as the second ratio. The first ratio of curve A3 is calculated. The product of the reciprocal of the median value within the horizontal axis range of the first rising trend curve segment and the second ratio is taken as the first value of curve A3. The data unreliability of the rising trend curve segment ; by As the curve A3 The credibility of each data point in the rising trend curve segment; is a linear normalization function; The credibility of each data point on curve A3 that is not in the rising trend curve segment is set to a preset second threshold.

9. The automatic testing and sorting method for chip MTF value according to claim 1, characterized in that: The specific steps of determining the optimal MTF value of the chip to be tested are as follows: Among all the brightness of light sources, the brightness of light source corresponding to the greatest brightness is taken as the optimal brightness of light source for the chip to be tested. The MTF value of the chip under test at a preset specific spatial frequency in all MTF charts under the optimal light source brightness is obtained, and the average of the MTF values ​​at the preset specific spatial frequency in all the MTF charts is taken as the optimal MTF value of the chip under test.

10. The automatic testing and sorting method of chip MTF value according to claim 1, characterized in that: The specific steps of sorting the chips to be tested are as follows: When the optimal MTF value is greater than the preset qualified threshold, the chip under test is determined to be qualified; When the optimal MTF value is less than or equal to the preset qualified threshold, the chip under test is determined to be unqualified.

Images