Image out-of-focus detection method, image out-of-focus detection system and critical dimension measurement method

Abstract

The invention provides an image out-of-focus detection method, a detection system thereof and a key dimension measurement method. According to the detection method, the characteristic that the boundary of a to-be-detected pattern in an image is blurred when the image is out of focus, so that the measurement result of the to-be-detected pattern by measurement equipment has deviation is utilized, and the difference value of the measurement widths of the obtained to-be-detected pattern at different positions is compared with a first standard value; judging whether the current detection image has an out-of-focus phenomenon or not; or the slope value of the linear relation can be obtained according to the measurement widths of the to-be-measured pattern at different positions, so that whether the detected image has an out-of-focus phenomenon or not can be judged according to whether the slope value exceeds the range of the corresponding second standard value or not. According to the invention, whether the currently acquired detection image has an out-of-focus phenomenon can be judged, and the problem of inaccurate measurement result caused by measurement of the key size based on the out-of-focus detection image is avoided.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor technology, and in particular to a method for detecting image defocus and a detection system thereof, and a method for measuring key dimensions. Background Art

[0002] With the continuous advancement of ultra-large-scale integrated circuit process technology, the critical dimension (CD) of semiconductor devices is getting smaller and smaller, and the process window is getting smaller and smaller. The measurement requirements for critical dimensions (for example, width dimensions) are also getting higher and higher to ensure the accuracy of critical dimension measurement. When measuring, generally after obtaining an image containing the pattern to be measured, the size of the pattern to be measured is measured based on the image. However, in the actual measurement process, the measuring equipment is prone to defocus when acquiring the image, which leads to abnormal measurement results. Summary of the invention

[0003] The object of the present invention is to provide a method and system for detecting image defocus, so as to solve the problem that the acquired detection image is not discovered in time when defocus occurs, resulting in abnormal measurement results.

[0004] In this regard, the present invention provides a method for detecting image defocus, comprising: first, obtaining a detection image containing a pattern to be measured, and obtaining the measured widths of the pattern to be measured in the detection image at at least two different positions; then, obtaining the difference between the measured widths at different positions, and comparing the difference with a corresponding first standard value, and when the difference exceeds the range of the corresponding first standard value, it is determined that the detection image is defocused; and / or, obtaining a linear relationship based on the measured widths of the pattern to be measured at at least two different positions, and obtaining the slope value of the linear relationship, and comparing the slope value with a corresponding second standard value, and when the slope value exceeds the range of the corresponding second standard value, it is determined that the detection image is defocused.

[0005] Optionally, the measured width of the pattern to be tested in the detection image at three or more different positions is obtained, and the difference in the measured width between multiple different positions is obtained, and each difference is compared with a corresponding first standard value; when any of the differences exceeds the range of the corresponding first standard value, it is determined that the detection image is out of focus.

[0006] Optionally, the method for setting the first standard value includes: obtaining theoretical widths of the pattern to be measured at at least two corresponding positions, and obtaining a theoretical difference between the theoretical widths at different positions to form the first standard value.

[0007] Optionally, the method for setting the first standard value includes: collecting historical measurement data of the pattern to be measured at at least two corresponding positions to obtain multiple measured widths of the pattern to be measured at each position, obtaining corresponding multiple groups of difference values ​​based on the multiple measured widths, and determining the range of the first standard value with the maximum difference and the minimum difference in the multiple groups of difference values.

[0008] Optionally, the method for setting the second standard value includes: obtaining the theoretical width of the pattern to be tested at least at two corresponding positions, and obtaining a corresponding linear relationship based on the theoretical widths of at least two corresponding positions to obtain a theoretical slope value for forming the second standard value.

[0009] Optionally, the method for setting the second standard value includes: collecting historical measurement data of the pattern to be measured at at least two corresponding positions to obtain multiple measured widths of the pattern to be measured at each position, obtaining corresponding multiple sets of linear relationship equations based on the multiple measured widths, and obtaining multiple slope values, and determining the range of the second standard value with the maximum slope value and the minimum slope value among the multiple slope values.

[0010] Optionally, the detection image displays a top view structure of the pattern to be detected, the top view structure is a strip-shaped figure extending along a predetermined direction, and the width of the top view structure is measured at different positions in the extending direction to obtain the measured width. Alternatively, the detection image displays a cross-sectional structure of the pattern to be detected along the height direction, and the width of the cross-sectional structure is measured at different height positions to obtain the measured width.

[0011] Optionally, the method for obtaining a linear relationship based on the measured width of the pattern to be measured at at least two different positions includes: establishing a rectangular coordinate system with the measured width direction of the pattern to be measured as the first axis and the direction perpendicular to the measured width as the second axis, and obtaining a corresponding linear relationship based on the rectangular coordinate system, one of the first axis and the second axis is the X-axis, and the other is the Y-axis.

[0012] Optionally, the method for obtaining a linear relationship based on the measured widths of the pattern to be measured at at least two different positions includes: establishing a rectangular coordinate system with the measured width direction of the pattern to be measured as the X-axis, the center line direction of the measured width as the Y-axis, and the center point of the measured width of the pattern to be measured at any position as the origin; obtaining a linear relationship y=kx+b in the rectangular coordinate system based on the measured widths of the pattern to be measured at at least two different positions; wherein y represents the distance of the measured position of the pattern to be measured relative to the origin position; x represents 1 / 2 of the measured width; k represents the slope of the boundary of the pattern to be measured; and b represents the distance of the virtual position relative to the origin position as the measured width is reduced to zero.

[0013] Optionally, the method of obtaining the linear relationship according to the measured widths of the pattern to be measured at least at two different positions includes: establishing a rectangular coordinate system with the measured width direction of the pattern to be measured as the Y axis, the center line direction of the measured width as the X axis, and the center point of the measured width of the pattern to be measured at any position as the origin;

[0014] According to the measured widths of the pattern to be measured at at least two different positions, a linear relationship y=kx+b is obtained in the rectangular coordinate system; wherein y represents 1 / 2 of the measured width; x represents the distance of the measured position of the pattern to be measured relative to the origin position; k represents the slope of the boundary of the pattern to be measured; and b represents 1 / 2 of the measured width at the origin position.

[0015] Optionally, before obtaining the linear relationship, it also includes: comparing the measured width of the pattern to be measured with the theoretical width of the pattern to be measured at the corresponding position, and determining whether the measured width is greater than the theoretical width; if so, determining that the pattern to be measured in the current detection image is tilted, adjusting the position of the pattern to be measured in the camera field of view, and reacquiring the detection image until the pattern to be measured in the detection image is not tilted; if not, determining that the pattern to be measured in the current detection image is not tilted, and obtaining the linear relationship based on the detection image.

[0016] Optionally, the method for detecting image defocus includes: obtaining a first measured width of a pattern to be measured in the detection image at a first position, a second measured width at a second position, and a third measured width at a third position; obtaining the difference between the first measured width and the second measured width, the difference between the first measured width and the third measured width, and the difference between the second measured width and the third measured width; when any of the differences exceeds the range of the corresponding first standard value, determining that the detection image is defocused; if all the differences do not exceed the range of the corresponding first standard value, executing the next step; obtaining a corresponding linear relationship based on the first measured width, the second measured width, and the third measured width, and obtaining a slope value of the linear relationship; and determining whether the slope value exceeds the range of the corresponding second standard value; if so, indicating that the detection image is defocused; if not, indicating that the detection image is not defocused.

[0017] The present invention provides a detection system for image defocus, comprising: an image acquisition unit, used for acquiring a detection image containing a pattern to be detected; a measuring unit, used for identifying the measurement boundary of the pattern to be detected in the detection image, and measuring the measurement width of the pattern to be detected at least at two different positions; and a data processing unit, used for obtaining the difference between the measurement widths of the pattern to be detected at different positions, and comparing the difference with a corresponding first standard value, and when the difference exceeds the range of the corresponding first standard value, judging that the detection image has a defocus phenomenon; and / or, the data processing unit is used for obtaining a linear relationship according to the measurement widths of the pattern to be detected at least at two different positions, and obtaining the slope value of the linear relationship, and comparing the slope value with a corresponding second standard value, and when the slope value exceeds the range of the corresponding second standard value, judging that the detection image has a defocus phenomenon.

[0018] Optionally, the measuring unit is used to measure the measured width of the pattern to be measured at three or more different positions; the data processing unit is used to obtain the difference in measured width between multiple different positions, and compare each difference with a corresponding first standard value. When any of the differences exceeds the range of the corresponding first standard value, it is determined that the detection image is out of focus.

[0019] Optionally, the data processing unit is used to establish a rectangular coordinate system with the measured width direction of the pattern to be measured as the first axis and the direction perpendicular to the measured width as the second axis when a linear relationship is obtained based on the measured width of the pattern to be measured at at least two different positions, and obtain a corresponding linear relationship based on the rectangular coordinate system, one of the first axis and the second axis is the X-axis, and the other is the Y-axis.

[0020] Optionally, the data processing unit is used to establish a rectangular coordinate system with the measured width direction of the pattern to be measured as the X-axis, the center line direction of the measured width as the Y-axis, and the center point of the measured width of the pattern to be measured at any position as the origin when a linear relationship is obtained based on the measured widths of the pattern to be measured at at least two different positions; or, establish a rectangular coordinate system with the measured width direction of the pattern to be measured as the Y-axis, the center line direction of the measured width as the X-axis, and the center point of the measured width of the pattern to be measured at any position as the origin.

[0021] Optionally, the data processing unit is also used to compare the measured width of the pattern to be measured with the theoretical width of the pattern to be measured at the corresponding position before obtaining the linear relationship, and determine whether the measured width is greater than the theoretical width; if so, adjust the position of the pattern to be measured in the camera field of view, and reacquire the detection image until the pattern to be measured in the detection image is not tilted; if not, obtain the linear relationship based on the detection image.

[0022] Optionally, the detection system further includes a storage unit for storing the first standard value and the second standard value.

[0023] The present invention also provides a method for measuring key dimensions, comprising: using the image defocus detection method as described above to determine whether the current detection image has a defocus phenomenon; if so, reacquiring the detection image and determining whether the acquired detection image has a defocus phenomenon, until the acquired detection image has no defocus phenomenon; and, performing key dimension measurement based on the detection image without a defocus phenomenon.

[0024] The present invention also provides a measuring device, comprising: the image defocusing detection system as described above; the measuring unit is also used to measure the key dimensions of the detection image without defocusing phenomenon.

[0025] In the image defocus detection method provided by the present invention, the characteristic that when the image is out of focus, the boundary of the pattern to be measured therein is blurred, so that the measuring device will have uncertainty when capturing the pattern boundary, resulting in an uncertain deviation in the measured measurement width (for example, at a certain position, the inner side of the blurred boundary is identified as the measurement boundary, resulting in a smaller measured width; and at another position, the outer side of the blurred boundary is identified as the measurement boundary, resulting in a larger measured width). The difference in the measured width of the pattern to be measured at different positions is compared with the first standard value to determine whether the current detection image has a defocus phenomenon; and / or, the slope value of a linear relationship is obtained based on the measured width of the pattern to be measured at different positions, so that it can be determined whether the detection image has a defocus phenomenon based on whether the slope value exceeds the range of the corresponding second standard value.

[0026] In the present invention, it is possible to judge whether the currently acquired detection image has a defocusing phenomenon, so when the detection image has a defocusing phenomenon, it can be discovered in time, avoiding the problem of inaccurate measurement results caused by measuring the key dimensions based on the defocused detection image. In other words, before measuring the key dimensions, the present invention preferentially judges whether the acquired detection image has a defocusing phenomenon, so that the measurement of the key dimensions is based on the detection image without defocusing, ensuring the accuracy of the measurement results. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 FIG. 4 is a flow chart of a method for detecting image defocus in one embodiment of the present invention.

[0028] Figure 2 FIG. 4 is a schematic diagram of a cross-sectional structure of a pattern to be tested in one embodiment of the present invention.

[0029] Figure 3FIG. 4 is a schematic diagram of a top view structure of a pattern to be tested in one embodiment of the present invention.

[0030] Figure 4 This is a first schematic diagram of establishing a rectangular coordinate system and obtaining a corresponding linear relationship according to the measured widths of the pattern to be measured at least at two different positions in one embodiment of the present invention.

[0031] Figure 5 This is a second schematic diagram of establishing a rectangular coordinate system and obtaining a corresponding linear relationship according to the measured widths of the pattern to be measured at least at two different positions in one embodiment of the present invention.

[0032] Figure 6 This is a third schematic diagram of establishing a rectangular coordinate system and obtaining a corresponding linear relationship according to the measured widths of the pattern to be measured at least at two different positions in one embodiment of the present invention.

[0033] Figure 7 This is a comparison chart of the measured width and the actual width when the pattern to be tested in the test image is tilted.

[0034] Figure 8 The figure is a flowchart of a specific detection process of a method for detecting image defocus in one embodiment of the present invention.

[0035] Fig. 9 FIG. 4 is a schematic diagram of a measuring device including an image defocus detection system in one embodiment of the present invention. DETAILED DESCRIPTION

[0036] As described in the background art, during the measurement of key dimensions, the measuring device is prone to defocusing when acquiring images, which in turn leads to abnormal measurement results.

[0037] In this regard, the inventor of the present invention has carefully studied the measurement state of the pattern when the image is out of focus by the measuring device, and found that when the image is out of focus, due to the blurred boundary of the pattern to be measured, the measuring device has uncertainty when capturing the pattern boundary, including: there are differences in the boundary recognition of different positions of the pattern to be measured, for example, for a certain position of the pattern to be measured, the outermost part of the blurred boundary is recognized as the measurement boundary, while for another position of the image to be measured, the innermost part of the blurred boundary is recognized as the measurement boundary. That is, when capturing the pattern boundary, the measuring device may recognize the outermost part of the blurred boundary as the measurement boundary, or recognize the innermost part of the blurred boundary as the measurement boundary, or recognize any position between the innermost and outermost parts of the blurred boundary as the measurement boundary.

[0038] In view of this, the present invention provides a method for detecting image defocus, which can be specifically referred to Figure 1 As shown, the detection method includes.

[0039] Step S10, obtaining a detection image containing a pattern to be detected, and obtaining, according to the detection image, the measured widths of the pattern to be detected in the detection image at least at two different positions.

[0040] Step S20, obtaining the difference between the measured widths at different positions, and comparing the difference with a first standard value. When the difference exceeds the range of the first standard value, it is determined that the detection image is out of focus; and / or, obtaining a linear relationship based on the measured widths of the pattern to be measured at at least two different positions, and obtaining the slope value of the linear relationship, and comparing the slope value with a corresponding second standard value. When the slope value exceeds the range of the corresponding second standard value, it is determined that the detection image is out of focus.

[0041] That is, in the image defocus detection method provided by the present invention, the boundary of the pattern to be measured will be blurred when the image is out of focus, so that there will be uncertainty in the measurement device when capturing the pattern boundary, resulting in uncertain deviation of the obtained measurement width.

[0042] The present invention utilizes this characteristic when the image is out of focus, and compares the difference in the measured width of the pattern to be measured at different positions with the first standard value to determine whether the current detection image is out of focus. Specifically, when the measurement device is measuring, when the obtained detection image is out of focus and the boundary of the pattern to be measured is blurred, the measurement device will have differences in the recognition of boundaries at different positions (for example, at a certain position, the inner side of the blurred boundary is recognized as the measurement boundary, which will cause the obtained measurement width to be smaller; and at another position, the outer side of the blurred boundary is recognized as the measurement boundary, which will cause the obtained measurement width to be larger), and at this time, the obtained difference will exceed the first standard value. Therefore, when the difference in the measured width of the pattern to be measured at different positions exceeds the range of the first standard value, it can be determined that the detection image is out of focus.

[0043] The present invention utilizes this characteristic when the image is out of focus, and can also obtain a linear relationship based on the measured width of the pattern to be measured at different positions, and obtain the corresponding slope value, so as to judge whether the detection image is out of focus according to whether the slope value exceeds the range of the corresponding second standard value. Specifically, when the measuring device is measuring, when the detection image obtained is out of focus and the boundary of the pattern to be measured is blurred, the measurement device will have differences in identifying the boundaries at different positions, and the obtained measured width will also deviate, and the slope value of the linear relationship obtained will also exceed the range of the second standard value. Therefore, when the obtained slope value exceeds the range of the second standard value, it can be judged that the detection image is out of focus.

[0044] In one example, the difference in measured widths at different positions can be preferentially obtained, and it can be determined whether the difference exceeds the range of the corresponding first standard value. When the difference exceeds the range of the corresponding first standard value, it can be determined that the current detection image is out of focus. If the difference does not exceed the range of the corresponding first standard value, a linear relationship can be obtained based on the measured widths of the pattern to be measured at at least two different positions, and the slope value of the linear relationship can be obtained. Then, it can be determined whether the slope value exceeds the range of the corresponding second standard value. If so, it can be determined that the current detection image is out of focus.

[0045] In another example, a linear relationship can be preferentially obtained based on the measured widths of the pattern to be tested at at least two different positions, and the slope value of the linear relationship can be obtained. When the slope value exceeds the range of the corresponding second standard value, it is determined that the current detection image is out of focus. If the slope value does not exceed the range of the corresponding second standard value, the difference in the measured widths at different positions can be further obtained, and then it is determined whether the difference exceeds the range of the corresponding first standard value. If so, it is determined that the current detection image is out of focus.

[0046] The following is a further detailed description of the image defocus detection method and its detection system, and the critical dimension measurement method proposed in the present invention in conjunction with the accompanying drawings and specific embodiments. According to the following description, the advantages and features of the present invention will become clearer. It should be noted that the drawings are all in a very simplified form and use non-precise proportions, which are only used to conveniently and clearly assist in explaining the purpose of the embodiments of the present invention. It should be recognized that relative terms such as "above", "below", "top", "bottom", "above" and "below" shown in the drawings can be used to describe the relationship between various elements. These relative terms are intended to cover different orientations of elements other than the orientations depicted in the drawings. For example, if the device is inverted relative to the view in the drawing, an element described as being "above" another element will now be below the element.

[0047] In a specific example, the pattern to be measured can be formed on a substrate. Also, a detection image containing the pattern to be measured can be obtained by using an image acquisition unit of the measuring device, wherein the acquired detection image can display, for example, a top view structure of the pattern to be measured, and the top view structure is, for example, a strip-shaped pattern extending along a predetermined direction (for example, Figure 3 Alternatively, the acquired detection image may display, for example, a cross-sectional structure of the pattern to be detected along the height direction, and its cross-sectional structure may be, for example, Figure 2 shown.

[0048] Then, based on the detection image, the measured width of the pattern to be detected in the detection image at least at two different positions can be obtained, and the difference between the measured widths at different positions can be obtained. Figure 3For the detection image shown in FIG. 1 , the width of the pattern to be detected can be measured at different positions in its extension direction; and based on Figure 2 For the inspection image shown in FIG. 1 , the width of the pattern to be inspected can be measured at different height positions.

[0049] Wherein, obtaining the measured width of the pattern to be tested in the detection image at least at two different positions specifically includes: obtaining the first measured width CD1 of the pattern to be tested in the detection image at the first position and the second measured width CD2 at the second position based on the obtained detection image, and obtaining the difference (CD1-CD2) between the first measured width and the second measured width. Afterwards, the obtained difference (CD1-CD2) can be compared with the corresponding first standard value Spec1 to determine whether the detection image is out of focus.

[0050] In a further solution, the measured width of the pattern to be tested at three or more different positions can be obtained, and then the difference of the measured width between the multiple different positions can be obtained, and each difference is compared with the corresponding first standard value. When any of the differences exceeds the range of the corresponding first standard value, it is determined that the obtained detection image is out of focus. Figure 2 and Figure 3 As shown, based on the obtained detection image, the first measurement width CD1 of the pattern to be measured in the detection image at the first position, the second measurement width CD2 at the second position, and the third measurement width CD3 at the third position are obtained, and the difference (CD1-CD2) between the first measurement width CD1 and the second measurement width CD2, the difference (CD1-CD3) between the first measurement width CD1 and the third measurement width CD3, and the difference (CD2-CD3) between the second measurement width CD2 and the third measurement width CD3 are obtained. When any of the differences exceeds the range of the corresponding first standard value Spec1, it is determined that the detection pattern is out of focus.

[0051] The first standard value Spec1 may be a threshold value range obtained by a theoretical difference △CD of theoretical widths of the pattern to be measured at different corresponding positions, and within an allowable error range, that is, Spec1 = △CD ± d, where d is an allowable error. Therefore, the method for setting the first standard value Spec1 may include: obtaining the theoretical widths of the pattern to be measured at least at two corresponding positions, and obtaining the theoretical difference △CD between the theoretical widths at different positions, for forming the first standard value Spec1.

[0052] In a specific example, combined with Figure 2 and Figure 3As shown, the first measurement width CD1 of the pattern to be measured at the first position, the second measurement width CD2 at the second position, and the third measurement width CD3 at the third position are obtained, and the difference (CD1-CD2) between the first measurement width CD1 and the second measurement width CD2, the difference (CD1-CD3) between the first measurement width CD1 and the third measurement width CD3, and the difference (CD2-CD3) between the second measurement width CD2 and the third measurement width CD3 are obtained; at this time, the theoretical difference △CD between the theoretical width of the pattern to be measured at the first position and the theoretical width of the pattern to be measured at the second position can be obtained accordingly. 1-2 , and the threshold range of the first standard value for the first position and the second position is obtained within the allowable error range (Spec1 = △CD 1-2 ±d); obtain the theoretical difference △CD between the theoretical width of the pattern to be measured at the first position and the theoretical width of the pattern to be measured at the third position 1-3 , and the threshold range of the first standard value for the first position and the third position is obtained within the allowable error range (Spec1 = △CD 1-3 ±d); obtain the theoretical difference △CD between the theoretical width of the pattern to be measured at the second position and the theoretical width of the pattern to be measured at the third position 2-3 , and obtain the threshold range of the first standard value for the second position and the third position within the allowable error range (Spec1 = △CD 2-3 ±d). After that, each difference value can be compared with the corresponding first standard value Spec1. When any difference value exceeds the range of the corresponding first standard value Spec1, it is determined that the obtained detection image is out of focus.

[0053] Alternatively, the first standard value Spec1 may also be obtained by collecting historical measurement data of the pattern to be measured, wherein the collected historical measurement data is specifically data obtained by measuring the pattern without losing focus, including multiple measurement widths of the pattern to be measured at various positions, and obtaining multiple corresponding sets of difference values ​​according to the multiple measurement widths of the pattern to be measured at different positions, and taking the maximum difference value △CD among the multiple sets of difference values ​​as the value. MAX and the minimum difference △CD MIN Determine the threshold range of the first standard value Spec1, that is, Spec1 = (△CD MAX , △CD MIN ). Therefore, the method for setting the first standard value Spec1 includes: obtaining historical measurement data of the pattern to be measured at at least two corresponding positions, obtaining multiple measurement widths of the pattern to be measured at each position, so as to obtain multiple sets of difference values ​​of the multiple measurement widths of the pattern to be measured at different positions, and taking the maximum difference value △CD among the multiple sets of difference values ​​as the difference value. MAX and the minimum difference △CD MINA threshold range of the first standard value Spec1 is determined.

[0054] For example, in a specific example, a first measured width CD1 of the pattern to be measured at the first position, a second measured width CD2 of the pattern to be measured at the second position, and a third measured width CD3 of the pattern to be measured at the third position are obtained, and the difference (CD1-CD2) between the first measured width CD1 and the second measured width CD2, the difference (CD1-CD3) between the first measured width CD1 and the third measured width CD3, and the difference (CD2-CD3) between the second measured width CD2 and the third measured width CD3 are obtained; at this time, historical measurement data of the pattern to be measured at the first position, the second position, and the third position can be obtained accordingly, and multiple groups of differences corresponding to multiple measured widths at the first position and the second position can be obtained (that is, the width of the pattern to be measured is measured multiple times at the first position and the second position in history, so that multiple groups of differences can be obtained), and the maximum difference △CD among the multiple groups of differences is used. MAX and the minimum difference △CD MIN The threshold range of the first standard value Spec1 for the first position and the second position is determined, that is, Spec1 = (△CD MAX-1-2 , △CD MIN-1-2 ); and obtaining multiple sets of difference values ​​corresponding to the multiple width measurements at the first position and the third position (that is, the width of the pattern to be measured is measured multiple times at the first position and the third position in history, so that multiple sets of difference values ​​can be obtained), and taking the maximum difference value △CD among the multiple sets of difference values MAX and the minimum difference △CD MIN The threshold range of the first standard value Spec1 for the first position and the third position is determined, that is, Spec1 = (△CD MAX-1-3 , △CD MIN-1-3 ); and obtaining multiple sets of difference values ​​corresponding to the multiple width measurements at the second position and the third position (that is, the width of the pattern to be measured is measured multiple times at the second position and the third position in history, so that multiple sets of difference values ​​can be obtained), and taking the maximum difference value △CD among the multiple sets of difference values MAX and the minimum difference △CD MIN The threshold range of the first standard value Spec1 for the second position and the third position is determined, that is, Spec1 = (△CD MAX-2-3 , △CD MIN-2-3 ). After that, the comparison results of each difference value and the corresponding first standard value Spec1 can be obtained. When any difference value exceeds the corresponding first standard value Spec1, it is determined that the detection pattern has a defocusing phenomenon.

[0055] In addition, in the detection method of image defocus provided by the present embodiment, in addition to using the method as described above to perform the first judgment, it also includes a second judgment process, which specifically includes: according to the measured width of the pattern to be tested at least at two different positions, a linear relationship y=kx+b is obtained, in which the measured width direction of the pattern is used as the first axis and the direction perpendicular to the measured width is used as the second axis to establish a rectangular coordinate system, and based on the rectangular coordinate system, one of the first axis and the second axis is the X axis and the other is the Y axis. For example, a rectangular coordinate system is established with the measured width direction of the pattern as the X axis and the direction perpendicular to the measured width as the Y axis; or, a rectangular coordinate system is established with the measured width direction of the pattern as the Y axis and the direction perpendicular to the measured width as the X axis. Afterwards, the obtained slope value k is compared with the corresponding second standard value spec2. When the obtained slope value k exceeds the range of the second standard value spec2, it is determined that the detected image is out of focus.

[0056] The method of obtaining the linear relationship y=kx+b according to the measured widths of the pattern to be measured at least at two different positions as described above is described below by listing several examples.

[0057] In the first example, please refer to Figure 4 As shown, a rectangular coordinate system is established with the measured width direction of the pattern as the X-axis, the center line direction of the measured width as the Y-axis, and the center point of the measured width of the pattern to be measured at any position (for example, the third position) as the origin. On this basis, the first measured width CD1 of the pattern to be measured at the first position, the second measured width CD2 of the pattern to be measured at the second position, and the third measured width CD3 of the pattern to be measured at the third position are obtained to obtain a linear relationship y=kx+b, wherein y represents the distance of the measured position of the pattern to be measured relative to the origin position (for example, the third position), x represents 1 / 2 of the measured width CD, the slope value k can directly represent the slope of the boundary of the pattern to be measured, and b represents the distance of the virtual position relative to the origin position (for example, the third position) when the measured width is reduced to zero under the slope k; thereafter, the obtained slope value k is compared with the corresponding second standard value spec2. When the obtained slope value k exceeds the range of the second standard value spec2, it is determined that the detection image is out of focus. At this time, whether the detection image is out of focus can be determined based on whether the boundary slope of the pattern to be detected in the detection image matches the slope standard value (ie, the second standard value spec2).

[0058] In the second example, please refer to Figure 5 As shown, a rectangular coordinate system is established with the measured width direction of the pattern as the Y axis, the center line direction of the measured width as the X axis, and the center point of the measured width at any position of the pattern to be measured as the origin. Figure 5 In the example, a rectangular coordinate system is established with the center point of the measured width at the third position as the origin. On this basis, the first measured width CD1, the second measured width CD2 and the third measured width CD3 are obtained to obtain the corresponding linear relationship y=kx+b, where x represents the distance of the measured position of the pattern to be measured relative to the origin position (i.e., the third position), y represents 1 / 2 of the measured width CD, the slope value k can directly represent the slope of the boundary of the pattern to be measured, and b represents 1 / 2 of the measured width CD at the origin position (i.e., the third position). At this time, it can be determined whether the detection image is out of focus based on whether the boundary slope of the pattern to be measured in the detection image is consistent with the slope standard value (i.e., the second standard value spec2).

[0059] In the third example, please refer to Figure 6 As shown, a rectangular coordinate system is established with the measured width direction of the pattern as the Y axis, the center line direction of the measured width as the X axis, and the center point of the measured width of the pattern to be measured at the second position as the origin. On this basis, the first measured width CD1, the second measured width CD2, and the third measured width CD3 are obtained to obtain the corresponding linear relationship y=kx+b, wherein x represents the distance of the measured position of the pattern to be measured relative to the origin position (the second position), y represents 1 / 2 of the measured width CD, the slope value k can directly represent the slope of the boundary of the pattern to be measured, and b represents 1 / 2 of the measured width CD at the origin position (the second position). At this time, it can be determined whether the detection image has a defocusing phenomenon based on whether the boundary slope of the pattern to be measured in the detection image is consistent with the slope standard value (i.e., the second standard value spec2).

[0060] The method for setting the second standard value spec2 may include: obtaining the theoretical width of the pattern to be tested at least at two corresponding positions, and obtaining the corresponding linear relationship y=k'x+b according to the theoretical width of at least two corresponding positions to obtain the theoretical slope value k', and the threshold range of the second standard value spec2 may be obtained in combination with the allowable error range, that is, Spec2=k'±d, where d is the allowable error. For example, the setting process of the second standard value spec2 may include: obtaining the first theoretical width of the pattern to be tested at the first position, the second theoretical width at the second position, and the third theoretical width at the third position, and obtaining the corresponding linear relationship y=k'x+b according to the first theoretical width, the second theoretical width and the third theoretical width, and then obtaining the theoretical slope value k', and the threshold range of the second standard value spec2 may be obtained in combination with the allowable error range.

[0061] Alternatively, the second standard value spec2 may be set by collecting historical measurement data of the pattern to be tested and obtaining the value based on the historical measurement data. Specifically, the method includes: obtaining historical measurement data of at least two corresponding positions of the pattern to be tested to obtain multiple measurement widths of the pattern to be tested at each position, and then obtaining multiple sets of corresponding linear relationship equations based on the multiple measurement widths, thereby obtaining multiple slope values ​​k, and taking the maximum slope value k among the multiple slope values ​​as the value. MAX and the minimum slope value k MIN Determine the threshold range of the second standard value Spec2, that is, Spec2 = (k MIN , k MAX ).

[0062] In an optional solution, before obtaining the linear relationship corresponding to the pattern to be tested and performing the second judgment process based on the corresponding slope value k, it may also include: judging whether the pattern to be tested in the current detection image is tilted. Specifically, for example, referring to Figure 7 As shown in FIG. 1 , when the pattern to be tested in the detection image is tilted, the measured width will be larger than the actual width. Figure 7 As shown, the measured width CD2' of the pattern to be tested at the second position is larger than the actual width CD2. Therefore, the method for judging whether the pattern to be tested in the current detection image is tilted may specifically include: comparing the measured width of the pattern to be tested with the actual width of the pattern to be tested at the corresponding position, and judging whether the measured width is greater than the actual width. If so, it is judged that the pattern to be tested in the current detection image is tilted. At this time, the position of the pattern to be tested in the camera field of view may be adjusted, and the detection image may be re-acquired until the pattern to be tested in the detection image is not tilted; if not, it is judged that the pattern to be tested in the current detection image is not tilted. On this basis, a corresponding linear relationship may be obtained based on the measured width of the pattern to be tested at different positions, and a subsequent second judgment process may be performed.

[0063] In order to more clearly explain the image defocus detection method provided in this embodiment, the following is combined with Figure 8 And for Figure 2 The specific detection process of the pattern to be tested shown is taken as an example for description.

[0064] First, step S100 is performed to obtain a detection image containing a pattern to be detected. Figure 2 As shown, the image to be tested displayed by the detection image is specifically a cross-sectional structure along the height direction, and the cross-sectional structure is, for example, in a trapezoidal shape.

[0065] Then, step S200 is performed to obtain the measured widths of the pattern to be tested in at least two different positions in the test image according to the test image. In this example, different height positions of the pattern to be tested are measured to obtain corresponding measured widths, that is, a first measured width CD1 of the pattern to be tested at a first position (e.g., the top position), a second measured width CD2 at a second position (e.g., the middle position), and a third measured width CD3 at a third position (e.g., the bottom position) are obtained.

[0066] Then, step S300 is performed to obtain the difference between the measured widths at different positions. In this example, it specifically includes: obtaining the difference (CD1-CD2) between the first measured width CD1 and the second measured width CD2, the difference (CD1-CD3) between the first measured width CD1 and the third measured width CD3, and the difference (CD2-CD3) between the second measured width CD2 and the third measured width CD3.

[0067] Then, step S400 is performed to compare the obtained difference with the corresponding first standard value spec1 to determine whether the difference exceeds the range of the corresponding first standard value spec1. In this example, the difference (CD1-CD2) between the first measurement width CD1 and the second measurement width CD2 is compared with the corresponding first standard value spec1 (△CD MAX-1-2 , △CD MIN-1-2 ) for comparison; comparing the difference (CD1-CD3) between the first measured width CD1 and the third measured width CD3 with the corresponding first standard value spec1 (△CD MAX-1-3 , △CD MIN-1-3 ) is compared, and the difference (CD2-CD3) between the second measured width CD2 and the third measured width CD3 is compared with the corresponding first standard value spec1 (△CD MAX-2-3 , △CD MIN-2-3 ) for comparison.

[0068] When any difference exceeds the range of the corresponding first standard value spec1, it is determined that the detection image is out of focus; otherwise, when all the difference values ​​do not exceed the range of the corresponding first standard value spec1, step S500 is continued.

[0069] At this point, the first judgment process of the detection image is completed, and the subsequent second judgment process can be continued. In a specific example, before the second judgment process is performed, it also includes: judging whether the pattern to be tested in the current detection image is tilted. The judgment method of whether the pattern to be tested is tilted can refer to the above description, which will not be repeated here. Among them, if the pattern to be tested in the current detection image is tilted, the position of the pattern to be tested in the camera field of view can be adjusted, and the detection image can be reacquired; if the pattern to be tested in the current detection image is not tilted, then on this basis, step S500 can be continued to perform the subsequent second judgment process.

[0070] Step S500 is executed to obtain a corresponding linear relationship according to the first measurement width CD1 , the second measurement width CD2 , and the third measurement width CD3 , and acquire a slope value k of the linear relationship.

[0071] Step S600 is executed to compare the obtained slope value k with the corresponding second standard value spec2 to determine whether the slope value k exceeds the range of the corresponding second standard value spec2. In this example, the second standard value Spec2 is the slope standard value, and its threshold range is, for example, Spec2 = (k MIN , k MAX ).

[0072] Among them, when the obtained slope value k exceeds the range of the corresponding second standard value spec2, it is judged that the detection image is out of focus; conversely, when the obtained slope value k does not exceed the range of the corresponding second standard value spec2, it is judged that the detection image does not have a defocus problem. At this time, based on the current detection image, the true boundary of the pattern to be tested can be accurately identified.

[0073] It should be noted that in Figure 8 In the example shown, the deviation of the difference of the measured widths at different positions relative to the first standard value is preferentially used as the first judgment; then a linear relationship is obtained based on the measured widths of the pattern to be measured at least at two different positions, and the deviation of the slope value in the linear relationship relative to the second standard value is used as the second judgment. However, in other examples, a linear relationship can also be preferentially obtained based on the measured widths of the pattern to be measured at least at two different positions, and the deviation of the slope value in the linear relationship relative to the second standard value is used as the first judgment; then, the deviation of the difference of the measured widths at different positions relative to the first standard value is used as the second judgment. Alternatively, in other examples, the first and second judgments as described above can also be performed simultaneously.

[0074] Based on the above-mentioned image defocus detection method, this embodiment also provides an image defocus detection system, which can be specifically referred to Fig. 9The schematic diagram of the measuring device containing the detection system shown in the figure, the detection system comprises: an image acquisition unit, used to acquire a detection image containing a pattern to be detected; a measuring unit, used to identify the measurement boundary of the pattern to be detected in the detection image (it should be noted that when the detection image is out of focus, the measurement boundary identified by the measuring unit may be biased and cannot represent the real boundary of the pattern to be detected; conversely, when the detection image is not out of focus, the measurement boundary identified by the measuring unit will be closer to the real boundary of the pattern to be detected), and measure the measurement width of the pattern to be detected at least at two different positions; and , a data processing unit is used to obtain the difference between the measured widths of the pattern to be tested at different positions, and compare the difference with the corresponding first standard value, and when the difference exceeds the range of the corresponding first standard value, it is judged that the detection image is out of focus; and / or, the data processing unit is used to obtain a linear relationship based on the measured widths of the pattern to be tested at least at two different positions, and obtain the slope value of the linear relationship, compare the slope value with the corresponding second standard value, and when the slope value exceeds the range of the corresponding second standard value, it is judged that the detection image is out of focus.

[0075] In a specific example, during the detection process of image defocusing of the detection system, the measuring unit can be used to measure the measured width of the pattern to be tested at three or more different positions, and the data processing unit is used to obtain the difference in the measured widths between multiple different positions, and compare each difference with the corresponding first standard value. When any of the differences exceeds the range of the corresponding first standard value, it is determined that the detected image is out of focus.

[0076] Further, when the data processing unit is used to obtain a linear relationship based on the measured width of the pattern to be measured at least at two different positions, a rectangular coordinate system is established with the measured width direction of the pattern to be measured as the first axis and the direction perpendicular to the measured width as the second axis, and a corresponding linear relationship is obtained based on the rectangular coordinate system, one of the first axis and the second axis is the X axis, and the other is the Y axis. For example, the data processing unit can specifically establish a rectangular coordinate system with the measured width direction of the pattern to be measured as the X axis, the center line direction of the measured width as the Y axis, and the center point of the measured width of the pattern to be measured at any position as the origin; or, establish a rectangular coordinate system with the measured width direction of the pattern to be measured as the Y axis, the center line direction of the measured width as the X axis, and the center point of the measured width of the pattern to be measured at any position as the origin.

[0077] In an optional example, the data processing unit is further used to compare the measured width of the pattern to be measured with the theoretical width of the pattern to be measured at the corresponding position before obtaining the linear relationship, and to determine whether the measured width is greater than the theoretical width; if so, adjust the position of the pattern to be measured within the camera field of view, and reacquire the detection image until there is no tilt in the pattern to be measured in the detection image; if not, obtain the linear relationship based on the detection image.

[0078] In a further embodiment, the detection system further includes a storage unit for storing the first standard value and the second standard value. Specifically, after obtaining the difference between the measured widths of the pattern to be tested at different positions, the data processing unit is further used to retrieve the corresponding first standard value from the storage unit for comparison processing; and / or, after obtaining the slope value of the corresponding linear relationship according to the measured widths of the pattern to be tested at least at two different positions, the data processing unit is further used to retrieve the corresponding second standard value from the storage unit for comparison processing.

[0079] Based on the above-mentioned image defocus detection method and detection system, this embodiment also provides a method for measuring key dimensions, which specifically includes: using the above-mentioned image defocus detection method to determine whether the current detection image has a defocus phenomenon; if so, reacquire the detection image, and determine whether the reacquired detection image has a defocus phenomenon, until the acquired detection image has no defocus phenomenon; then, the key dimension can be measured based on the detection image without defocus phenomenon. In this embodiment, before measuring the key dimension, it is first determined whether the acquired detection image has a defocus phenomenon, so as to ensure that the measurement of the key dimension is based on the detection image without defocus phenomenon, so as to ensure the accuracy of the measurement result.

[0080] In addition, this embodiment also provides a measuring device, such as Fig. 9 As shown, it specifically includes: the image defocus detection system as described above, which is used to determine whether the current detection image has a defocus phenomenon; and, when the acquired detection image does not have a defocus phenomenon, the measurement unit is also used to measure the key dimensions of the detection image without a defocus phenomenon. Wherein, the measuring device can be a scanning electron microscope (SEM). In a specific example, the measuring device also includes a feedback unit, which is used to receive the detection result of whether the current detection image is defocused from the data processing unit, and when the detection result is that the current detection image has a defocus phenomenon, the feedback unit can send a command signal to the image acquisition unit to instruct the image acquisition unit to re-acquire the detection image; and, when the detection result is that the current detection image does not have a defocus phenomenon, the feedback unit can send a command signal to the measurement unit to instruct the measurement unit to perform key dimension measurement based on the current detection image without a defocus phenomenon.

[0081] It should be noted that, although the present invention has been disclosed as a preferred embodiment, the above embodiment is not intended to limit the present invention. For any technician familiar with the art, without departing from the scope of the technical solution of the present invention, the technical content disclosed above can be used to make many possible changes and modifications to the technical solution of the present invention, or modified into equivalent embodiments of equivalent changes. Therefore, any simple modification, equivalent change and modification made to the above embodiment according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still belongs to the scope of protection of the technical solution of the present invention.

[0082] It must be noted that as used herein and in the appended claims, the singular forms "a," "an," and "an" include plural references unless the context clearly indicates otherwise. For example, reference to "a step" or "a device" means reference to one or more steps or devices and may include secondary steps as well as secondary devices. All conjunctions used should be interpreted in the broadest sense. Also, the word "or" should be interpreted as having the definition of a logical "or" rather than a logical "exclusive or" unless the context clearly indicates otherwise.

Claims

1. A method for detecting image defocus, characterized in that: include: Acquire a test image containing a pattern to be tested, and obtain measured widths of the pattern to be tested in the test image at least at two different positions; Obtaining the difference between the measured widths at different positions, and comparing the difference with the corresponding first standard value, and when the difference exceeds the range of the corresponding first standard value, determining that the detected image is out of focus; And / or, a linear relationship is obtained based on the measured width of the pattern to be tested at at least two different positions, and the slope value of the linear relationship is obtained, and the slope value is compared with a corresponding second standard value. When the slope value exceeds the range of the corresponding second standard value, it is determined that the detection image is out of focus.

2. The method for detecting image defocus according to claim 1, characterized in that: The measured widths of the pattern to be tested in the detection image at three or more different positions are obtained, and the difference in the measured widths between the multiple different positions is obtained, and each difference is compared with a corresponding first standard value; when any of the differences exceeds the range of the corresponding first standard value, it is determined that the detection image is out of focus.

3. The method for detecting image defocus according to claim 1, characterized in that: The method for setting the first standard value includes: obtaining theoretical widths of the pattern to be tested at at least two corresponding positions, and obtaining theoretical differences between theoretical widths at different positions to form the first standard value; or, The method for setting the first standard value includes: collecting historical measurement data of the pattern to be measured at at least two corresponding positions to obtain multiple measured widths of the pattern to be measured at each position, obtaining corresponding multiple groups of difference values ​​based on the multiple measured widths, and determining the range of the first standard value with the maximum difference and the minimum difference in the multiple groups of difference values.

4. The method for detecting image defocus according to claim 1, characterized in that: The method for setting the second standard value includes: obtaining the theoretical width of the pattern to be tested at least at two corresponding positions, and obtaining a corresponding linear relationship according to the theoretical width of the at least two corresponding positions to obtain a theoretical slope value for forming the second standard value; or, The method for setting the second standard value includes: collecting historical measurement data of the pattern to be measured at at least two corresponding positions to obtain multiple measured widths of the pattern to be measured at each position, obtaining corresponding multiple sets of linear relationship equations based on the multiple measured widths, and obtaining multiple slope values, and determining the range of the second standard value with the maximum slope value and the minimum slope value among the multiple slope values.

5. The method for detecting image defocus according to claim 1, characterized in that: The detection image displays a top-view structure of the pattern to be detected, the top-view structure is a strip-shaped figure extending along a predetermined direction, and the width of the top-view structure is measured at different positions in the extending direction thereof to obtain the measured width; Alternatively, the detection image displays a cross-sectional structure of the pattern to be detected along a height direction, and the width of the cross-sectional structure is measured at different height positions to obtain the measured width.

6. The method for detecting image defocus according to claim 1, characterized in that: The method for obtaining a linear relationship based on the measured width of the pattern to be measured at at least two different positions includes: establishing a rectangular coordinate system with the measured width direction of the pattern to be measured as a first axis and a direction perpendicular to the measured width as a second axis, and obtaining a corresponding linear relationship based on the rectangular coordinate system, wherein one of the first axis and the second axis is an X-axis and the other is a Y-axis.

7. The method for detecting image defocus according to claim 6, characterized in that: The method for obtaining a linear relationship according to the measured widths of the pattern to be measured at least at two different positions includes: establishing a rectangular coordinate system with the measured width direction of the pattern to be measured as the X-axis, the center line direction of the measured width as the Y-axis, and the center point of the measured width of the pattern to be measured at any position as the origin; Obtaining a linear relationship y=kx+b in the rectangular coordinate system according to the measured widths of the pattern to be measured at least at two different positions; Wherein, y represents the distance of the measurement position of the pattern to be measured relative to the origin position; x represents 1 / 2 of the measured width; k represents the slope of the boundary of the pattern to be measured; b represents the distance of the virtual position relative to the origin position as the measurement width is reduced to zero.

8. The method for detecting image defocus according to claim 6, wherein: The method for obtaining a linear relationship according to the measured width of the pattern to be measured at least at two different positions comprises: establishing a rectangular coordinate system with the measured width direction of the pattern to be measured as the Y axis, the center line direction of the measured width as the X axis, and the center point of the measured width of the pattern to be measured at any position as the origin; Obtaining a linear relationship y=kx+b in the rectangular coordinate system according to the measured widths of the pattern to be measured at least at two different positions; Among them, y represents 1 / 2 of the measured width; x represents the distance of the measured position of the pattern to be measured relative to the origin position; k represents the slope of the boundary of the pattern to be measured; b represents 1 / 2 of the measured width at the origin.

9. The method for detecting image defocus according to any one of claims 1, 6 to 8, characterized in that: Before obtaining the linear relationship, the method further includes: Comparing the measured width of the pattern to be measured with the theoretical width of the pattern to be measured at the corresponding position, and determining whether the measured width is greater than the theoretical width; If yes, it is determined that the pattern to be tested in the current detection image is tilted, the position of the pattern to be tested in the camera field of view is adjusted, and the detection image is acquired again until the pattern to be tested in the detection image is no longer tilted; If not, it is determined that the pattern to be tested in the current test image does not have a tilt, and the linear relationship is obtained based on the test image.

10. The method for detecting image defocus according to claim 1, characterized in that: include: Acquire a first measured width of the pattern to be measured in the detection image at a first position, a second measured width at a second position, and a third measured width at a third position; Obtaining a difference between the first measured width and the second measured width, a difference between the first measured width and the third measured width, and a difference between the second measured width and the third measured width; When any of the difference values ​​exceeds the range of the corresponding first standard value, it is determined that the detection image is out of focus; if all the difference values ​​do not exceed the range of the corresponding first standard value, the next step is executed; Obtaining a corresponding linear relationship according to the first measured width, the second measured width, and the third measured width, and acquiring a slope value of the linear relationship; and, Determine whether the slope value exceeds the range of the corresponding second standard value; if so, it indicates that the detection image is out of focus; if not, it indicates that the detection image is not out of focus.

11. A system for detecting image defocus, characterized in that: include: An image acquisition unit, used for acquiring a detection image containing a pattern to be detected; A measuring unit, used to identify a measurement boundary of the pattern to be measured in the detection image, and measure a measurement width of the pattern to be measured at least at two different positions; as well as, A data processing unit, used for obtaining the difference between the measured widths of the pattern to be tested at different positions, and comparing the difference with the corresponding first standard value, and determining that the detected image is out of focus when the difference exceeds the range of the corresponding first standard value; And / or, a linear relationship is obtained based on the measured width of the pattern to be tested at at least two different positions, and the slope value of the linear relationship is obtained, and the slope value is compared with a corresponding second standard value. When the slope value exceeds the range of the corresponding second standard value, it is determined that the detection image is out of focus.

12. The image defocus detection system according to claim 11, characterized in that: The measuring unit is used to measure the measured width of the pattern to be measured at three or more different positions; the data processing unit is used to obtain the difference in measured width between multiple different positions, and compare each difference with the corresponding first standard value. When any of the differences exceeds the range of the corresponding first standard value, it is determined that the detection image is out of focus.

13. The image defocus detection system according to claim 11, characterized in that: The data processing unit is used to establish a rectangular coordinate system with the measured width direction of the pattern to be measured as the first axis and the direction perpendicular to the measured width as the second axis when obtaining a linear relationship based on the measured width of the pattern to be measured at at least two different positions, and obtain a corresponding linear relationship based on the rectangular coordinate system, one of the first axis and the second axis is the X axis, and the other is the Y axis.

14. The image defocus detection system according to claim 13, characterized in that: The data processing unit is used to establish a rectangular coordinate system with the measured width direction of the pattern to be measured as the X-axis, the center line direction of the measured width as the Y-axis, and the center point of the measured width of the pattern to be measured at any position as the origin when a linear relationship is obtained according to the measured width of the pattern to be measured at at least two different positions; or, establish a rectangular coordinate system with the measured width direction of the pattern to be measured as the Y-axis, the center line direction of the measured width as the X-axis, and the center point of the measured width of the pattern to be measured at any position as the origin.

15. The image defocus detection system according to any one of claims 11, 13-14, characterized in that: The data processing unit is also used to compare the measured width of the pattern to be measured with the theoretical width of the pattern to be measured at the corresponding position before obtaining the linear relationship, and determine whether the measured width is greater than the theoretical width; if so, adjust the position of the pattern to be measured in the camera field of view, and reacquire the detection image until the pattern to be measured in the detection image is not tilted; if not, obtain the linear relationship based on the detection image.

16. A method for measuring a critical dimension, characterized in that: include: Adopting the image defocus detection method as described in any one of claims 1 to 10 to determine whether the current detection image has a defocus phenomenon; If yes, reacquire the detection image, and determine whether the reacquired detection image has a defocusing phenomenon, until the reacquired detection image does not have a defocusing phenomenon; and, Critical dimension measurements based on inspection images without out-of-focus artifacts.

17. A measuring device, characterized in that include: The image defocus detection system as described in any one of claims 11-15; the measurement unit is also used to measure the key dimensions of the detection image without defocus phenomenon.

Images