Position detection method and system

Abstract

The invention provides a position detection method and system, and the method comprises the steps: arranging at least one detection mark on each chip of a substrate, scanning the whole substrate in a row / column direction, obtaining a plurality of substrate images containing at least one detection mark, recognizing the detection marks in the substrate images, and carrying out the recognition of the detection marks in the substrate images. And obtaining the actual pixel position of the detection mark in the substrate image, and obtaining the actual physical position of the detection mark according to the actual pixel position and the theoretical physical position of the detection mark so as to obtain the actual physical position of each chip. The detection marks are arranged on the chips, the actual physical position of each chip can be obtained by obtaining the actual physical position of the detection marks, the detection efficiency is high, and errors are small.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor technology, and in particular to a position detection method and system. Background Art

[0002] Wafer level chip-scale packaging (WLCSP) not only has the smallest area among all IC packaging forms, but also has excellent electrical and thermal performance. The WLCSP process can be divided into two types: direct BOP (bump on pad) and rewiring (RDL). BOP means that the solder balls grow directly on the chip's pads. Sometimes, if the pads for leading out the solder balls are too close and it is inconvenient to lead out the balls, RDL is used to lead the solder balls to the side. The current WLCSP process is Fan-out, that is, the chip is first cut off the wafer and inverted and glued to the substrate (Carrier). At this time, the substrate and the chip are bonded together to form a new wafer, called a reconstituted wafer, and RDL is exposed and grown in the reconstituted wafer.

[0003] In the Fan-out process, position errors may be introduced during the transfer of chips to the substrate, resulting in a certain difference between the actual position of each chip on the substrate and the theoretical position. How to detect the position of each chip on the substrate is a problem that needs to be solved urgently. Summary of the invention

[0004] The object of the present invention is to provide a position detection method and system, which can accurately detect the actual physical position of each chip on a substrate.

[0005] In order to achieve the above object, the present invention provides a position detection method, comprising:

[0006] Providing a substrate, wherein the substrate has a plurality of chips, and each of the chips has at least one detection mark;

[0007] Scanning the entire substrate along a row / column direction to obtain a plurality of substrate images including at least one detection mark;

[0008] identifying the detection mark in the substrate image, and obtaining the actual pixel position of the detection mark in the substrate image; and,

[0009] The actual physical position of the detection mark is acquired according to the actual pixel position and the theoretical physical position of the detection mark to obtain the actual physical position of each chip.

[0010] Optionally, the detection mark is identified in the substrate image using a template image of the detection mark through a template matching algorithm; and / or the detection mark is directly outlined in the substrate image.

[0011] Optionally, the template image is automatically extracted from any of the substrate images using an automatic marker extraction algorithm.

[0012] Optionally, the automatic label extraction algorithm includes at least one of an extraction algorithm based on image saliency, an extraction algorithm based on information entropy, and an extraction algorithm based on matrix principal components.

[0013] Optionally, the template image is cut out from any of the substrate images.

[0014] Optionally, the step of intercepting the template image includes:

[0015] Extracting a pre-processed image including the detection mark from any of the substrate images; and

[0016] The detection mark is identified in the pre-processed image, and the template image including the detection mark is cut out from the pre-processed image with the center of the identified detection mark as the center.

[0017] Optionally, before intercepting the template image, the position detection method further includes:

[0018] Selecting one of the chips on the substrate as a template chip, moving the field of view center of the scanning module to the theoretical physical positions of at least two vertices of the template chip one by one and taking photos to obtain at least two vertex images;

[0019] Obtaining the rotation angle deviation of the template chip relative to its theoretical physical position according to the pixel position deviation between the vertex in each vertex image and the image center; and

[0020] The substrate image corresponding to the template chip is rotated according to the rotation angle deviation, and the template chip in the rotated substrate image is selected for interception.

[0021] Optionally, before scanning the substrate, an image of the detection mark at the same magnification as that of a lens of a scanning module is acquired in advance as the template image.

[0022] Optionally, the template matching algorithm includes an ncc algorithm.

[0023] Optionally, before scanning the substrate, a mapping relationship between a coordinate system of the substrate and a coordinate system of a scanning module is acquired.

[0024] Optionally, the step of acquiring the actual physical position of the detection mark according to the actual pixel position and the theoretical physical position of the detection mark includes:

[0025] Acquiring a theoretical pixel position of the detection mark according to the theoretical physical position of the detection mark and the mapping relationship;

[0026] Obtaining a pixel position deviation between a theoretical pixel position and an actual pixel position of the detection mark;

[0027] converting the pixel position deviation of the detection mark into a physical position deviation according to the mapping relationship; and

[0028] The actual physical position of the detection mark is obtained according to the physical position deviation and the theoretical physical position of the detection mark.

[0029] Optionally, after acquiring the actual physical position of the detection mark, the position detection method further includes:

[0030] Acquiring a theoretical pixel position of the detection mark according to the theoretical physical position of the detection mark and the mapping relationship;

[0031] Determining whether the substrate image is stretched or compressed according to the theoretical pixel positions and actual pixel positions of all the detection marks in the substrate image; and

[0032] The actual physical position of the detection mark in the substrate image is corrected according to the stretching ratio or the compression ratio of the substrate image.

[0033] Optionally, the step of determining whether the substrate image is stretched or compressed includes:

[0034] Acquire an actual minimum circumscribed rectangle capable of enclosing all the detection marks in the substrate image according to actual pixel positions of all the detection marks in the substrate image;

[0035] Acquire a theoretical minimum circumscribed rectangle capable of enclosing all the detection marks in the substrate image according to theoretical pixel positions of all the detection marks in the substrate image; and

[0036] When the actual minimum circumscribed rectangle is equal to the theoretical minimum circumscribed rectangle in size, it is determined that the substrate image is not stretched or compressed; when the actual minimum circumscribed rectangle is not equal to the theoretical minimum circumscribed rectangle in size, it is determined that the substrate image is stretched or compressed.

[0037] Optionally, when the width W1 of the theoretical minimum circumscribed rectangle is not equal to the width W2 of the actual minimum circumscribed rectangle, it is determined that the substrate image is stretched or compressed in the width direction, and the stretching ratio or compression ratio is W1 / W2; when the height H1 of the theoretical minimum circumscribed rectangle is not equal to the height H2 of the actual minimum circumscribed rectangle, it is determined that the substrate image is stretched or compressed in the height direction, and the stretching ratio or compression ratio is H1 / H2.

[0038] Optionally, the actual physical position of the detection mark in the substrate image is assumed to be (x i ,y i ), wherein i is the serial number of the detection mark in the substrate image, the x direction is the width direction of the substrate image, and the y direction is the height direction of the substrate image; and,

[0039] When the substrate image is stretched or compressed in the width direction, the x coordinate of the detection mark i is corrected using the stretching ratio or compression ratio W1 / W2 of the substrate image. The corrected x coordinate is x i *(W1 / W2), when the substrate image is stretched or compressed in the height direction, the y coordinate of the detection mark i is corrected by the stretching ratio or compression ratio H1 / H2 of the substrate image, and the corrected y coordinate is y i *(H1 / H2).

[0040] Optionally, after acquiring the actual physical position of the detection mark, the position detection method further includes:

[0041] Selecting all the substrate images in any scanning queue, taking the first substrate image as a reference image, and obtaining the starting physical position deviations of the other substrate images in the scanning queue and the reference image one by one; and,

[0042] The actual physical positions of the detection marks in the corresponding substrate images in all the scanning queues are corrected according to the initial physical position deviation.

[0043] A group of substrate images with corresponding positions are selected from all scanning queues, and the substrate image in the first scanning queue is used as a reference image, and the starting physical position deviations of the substrate images in other scanning queues and the reference image are obtained one by one; and

[0044] The actual physical positions of the detection marks in all the substrate images in each scanning queue are corrected according to the starting physical position deviation of the substrate image in the scanning queue.

[0045] Optionally, the actual physical position of the detection mark in the substrate image is assumed to be (x jkn ,yjk ), wherein j is the sequence number of the scanning queue, k is the sequence number of the substrate image in the scanning queue, n is the sequence number of the detection mark in the substrate image, and y is the scanning direction;

[0046] According to the starting physical position deviation Δy of the substrate image of each scanning queue j Correct the y coordinates of the detection marks in all the substrate images in the scanning queue, and the corrected y coordinates are y jk +Δy j .

[0047] The present invention also provides a position detection system, comprising:

[0048] A motion module, used for carrying a substrate, the substrate having a plurality of chips, each of the chips having at least one detection mark;

[0049] A scanning module, used for scanning the entire substrate along a row / column direction to obtain a plurality of substrate images including at least one detection mark;

[0050] an identification module, used to identify the detection mark in the substrate image and obtain the actual pixel position of the detection mark in the substrate image; and

[0051] A position acquisition module is used to acquire the actual physical position of the detection mark according to the actual pixel position and theoretical physical position of the detection mark to obtain the actual physical position of each chip

[0052] In the position detection method and system provided by the present invention, at least one detection mark is set on each chip of the substrate, and then the entire substrate is scanned along the row / column direction to obtain a number of substrate images containing at least one detection mark, identify the detection mark in the substrate image, and obtain the actual pixel position of the detection mark in the substrate image, and obtain the actual physical position of the detection mark according to the actual pixel position and theoretical physical position of the detection mark to obtain the actual physical position of each chip. The present invention can obtain the actual physical position of each chip by setting the detection mark on the chip and obtaining the actual physical position of the detection mark, with high detection efficiency and small error. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A flow chart of a position detection method provided by an embodiment of the present invention;

[0054] Figure 2 A schematic structural diagram of a substrate provided in an embodiment of the present invention;

[0055] Figure 3An enlarged schematic diagram of a chip provided in an embodiment of the present invention;

[0056] Figure 4 A partial schematic diagram of a substrate provided in an embodiment of the present invention;

[0057] Figure 5 A schematic diagram of a substrate image provided by an embodiment of the present invention;

[0058] Figure 6a to Figure 6d A schematic diagram corresponding to the corresponding steps of manually intercepting the template image provided by an embodiment of the present invention;

[0059] Figure 7 A partial schematic diagram of a substrate provided in an embodiment of the present invention;

[0060] Figure 8 To include Figure 7 The substrate image of the first 6 detection marks in the first row;

[0061] Fig. 9 For Figure 8 Schematic diagram of extracting a template image from a substrate image in FIG.

[0062] Fig.10 A schematic diagram of a template chip provided by an embodiment of the present invention;

[0063] Fig.11 A vertex image obtained by photographing the vertices of a template chip provided in an embodiment of the present invention;

[0064] Fig.12 A vertex image obtained by photographing another vertex of the template chip provided by an embodiment of the present invention;

[0065] Fig.13 A schematic diagram of three scanning queues provided in an embodiment of the present invention;

[0066] Fig.14 A simplified schematic diagram of a position detection system provided by an embodiment of the present invention;

[0067] Wherein, the accompanying drawings are marked as follows:

[0068] 100 - substrate; 101 - chip; 101' - template chip; 111 - detection mark; 200 - substrate image; 201 - pre-processed image; 202 - template image; 301, 302 - vertex images; 10 - motion module; 20 - scanning module; 30 - recognition module; 40 - position acquisition module. DETAILED DESCRIPTION

[0069] The specific implementation of the present invention will be described in more detail below in conjunction with the schematic diagram. The advantages and features of the present invention will become clearer based on the following description. It should be noted that the drawings are all in a very simplified form and are not in exact proportions, and are only used to facilitate and clearly assist in explaining the purpose of the embodiments of the present invention.

[0070] Figure 1 Flow chart of the position detection method provided in this embodiment. Figure 1 As shown, the position detection method includes:

[0071] Step S100: providing a substrate, wherein the substrate has a plurality of chips, and each of the chips has at least one detection mark;

[0072] Step S200: scanning the entire substrate along the row / column direction to obtain a plurality of substrate images including at least one detection mark;

[0073] Step S300: identifying the detection mark in the substrate image, and obtaining the actual pixel position of the detection mark in the substrate image; and,

[0074] Step S400: Acquire the actual physical position of the detection mark according to the actual pixel position and the theoretical physical position of the detection mark to obtain the actual physical position of each chip.

[0075] See also Figure 2 to Figure 13 , which is a schematic diagram of the structure corresponding to the corresponding steps of the position detection method provided in this embodiment. Next, Figure 2 to Figure 13 The position detection method provided in this embodiment is described in detail.

[0076] Step S100 is executed to provide the substrate. Figure 2 The schematic diagram of the structure of the substrate 100 provided in this embodiment is as follows: Figure 2 As shown, the substrate 100 is a rectangle, and its size may be, for example, 510 mm*515 mm. The coordinate system of the substrate 100 is established with the short side of the substrate 100 as the X-axis and the long side as the Y-axis. The missing corner of the substrate 100 (for positioning) is at the lower left corner.

[0077] Please continue reading Figure 2 The substrate 100 has a plurality of chips 101 thereon. The chips 101 are cut from an original wafer and then invertedly glued to the substrate 100. Figure 2 FIG. 1 shows an ideal distribution of the chips 101 on the substrate 100. Figure 2 As can be seen in the figure, ideally, the chips 101 are distributed in an array on the substrate 100 .

[0078] Figure 3 FIG. 1 is an enlarged schematic diagram of the chip 101 provided in this embodiment. Figure 3 As shown, each of the chips 101 has four detection marks 111, and the four detection marks 111 are respectively arranged at the four corners of the chip 101, and each of the detection marks 111 has the same shape, which is a cross, and each of the detection marks 111 has the same size.

[0079] It should be understood that the chip 101 is not limited to being provided with 4 detection marks 111, and 1, 2, 3 or more than 4 detection marks 111 may be provided; the shape of each detection mark 111 may be different or partially the same; the size of each detection mark 111 may also be different or partially the same; examples will not be given one by one here.

[0080] Next, step S200 is performed to place the substrate 100 on a moving table with a scanning module above the moving table, and establish a mapping relationship between the coordinate system of the substrate 100 and the coordinate system of the scanning module. For example, the coordinate system of the substrate 100 can be the same as the coordinate system of the scanning module or a predetermined deviation can be set between the coordinate system of the substrate 100 and the coordinate system of the scanning module.

[0081] The process of establishing the mapping relationship between the coordinate system of the substrate 100 and the coordinate system of the scanning module can be to use the global mark ( Figure 2 (not shown) the substrate 100 is coarsely aligned and finely aligned. The method of coarsely aligning and finely aligning the substrate 100 can be any existing method, which will not be described in detail here.

[0082] Next, the entire substrate 100 is scanned along the y direction (row direction) to obtain a plurality of substrate images 200 including at least one detection mark 111. It should be understood that due to the size limitation of the field of view, the field of view of the scanning module is usually difficult to frame a whole row of the detection marks 111 on the substrate 100, and it is also difficult to frame two rows of the detection marks 111. Figure 4 A partial schematic diagram of the substrate 100 provided in this embodiment, Figure 5 A schematic diagram of a substrate image 200 provided in this embodiment, such as Figure 4 and Figure 5 As shown, in this embodiment, the field of view of the scanning module can frame six of the detection marks 111 , so each of the substrate images 200 includes six of the detection marks 111 .

[0083] Please continue reading Figure 4Before scanning the entire substrate 100, the scanning path of the scanning module can be planned according to the ideal distribution of the chip 101 on the substrate 100 and the theoretical physical position of the detection mark 111 on the chip 101. When scanning the substrate 100, the scanning module can move along the planned scanning path and photograph the detection mark 111 line by line. For example, the scanning module can first move to the initial position (for example, the lower left corner of the substrate 100) and set the field of view ( Figure 4 The scanning module 200 is aligned with the first six detection marks 111 in the first row (with the virtual box in the figure), and after the six detection marks 111 are located in the center of the field of view, the scanning module starts to take pictures to obtain the first substrate image 200; then, the scanning module moves along the y direction until the field of view is aligned with the first six detection marks 111 in the second row, and after the six detection marks 111 are located in the center of the field of view, the scanning module starts to take pictures to obtain the second substrate image 200; ... until the entire substrate is scanned and all the detection marks 111 are photographed, so that each detection mark 111 has a corresponding substrate image 200.

[0084] It should be noted that the substrate image 200 is not limited to containing 6 detection marks 111. The number of the detection marks 111 contained in the substrate image 200 depends on the field of view size of the scanning module. The scanning module is not limited to scanning the entire substrate 100 along the y direction, and can also scan the entire substrate 100 along the x direction (row direction), which does not affect the implementation of the present invention.

[0085] Next, step S300 is performed to identify all the detection marks 111 in each of the substrate images 200, thereby obtaining the actual pixel position of each of the detection marks 111 in the substrate image 200. In some embodiments, each of the substrate images 200 may be automatically identified by using a template matching algorithm using a template image of the detection mark 111, thereby identifying all of the detection marks 111 in each of the substrate images 200, or all of the detection marks 111 may be directly manually outlined in each of the substrate images 200, or a combination of the above two methods may be used to identify all of the detection marks 111 in each of the substrate images 200.

[0086] The template matching algorithm may be, for example, an ncc algorithm, which calculates the scores of individual pixels of the template image and the pixels in the substrate image 200 for matching, thereby identifying all the detection marks 111 in the substrate image 200 .

[0087] Since the template matching algorithm matches the center of the detection mark 111 in the substrate image 200 with the center of the template image, the position deviation of the detection mark 111 in the template image needs to be as small as possible, and the detection mark 111 in the template image should be located in the center of the template image as much as possible (the center of the detection mark 111 in the template image should coincide with the image center of the template image as much as possible), so as to reduce the matching error.

[0088] In some embodiments, the template image can be obtained online. For example, the template image can be automatically extracted from any of the substrate images 200 using an automatic label extraction algorithm, wherein the automatic label extraction algorithm includes at least one of an extraction algorithm based on image saliency, an extraction algorithm based on information entropy, and an extraction algorithm based on matrix principal components.

[0089] The template image may be obtained online by cutting out the template image from any of the substrate images 200 . Figure 6a to Figure 6d The following is a schematic diagram corresponding to the steps of manually capturing the template image provided in this embodiment. Figure 6a and Figure 6b As shown, first, a pre-processed image 201 containing the detection mark 111 is manually intercepted from any of the substrate images 200. Since it is manually intercepted, the detection mark 111 is very likely not at the center of the pre-processed image 201 (the center of the detection mark 111 in the pre-processed image 201 does not coincide with the image center of the pre-processed image 201). Figure 6c and Figure 6d As shown, the detection mark 111 is identified in the preprocessed image 201 by using an edge recognition algorithm, so that the center of the detection mark 111 in the preprocessed image 201 can be obtained, and then the template image 202 including the detection mark 111 is cut out in the preprocessed image 201 with the center of the detection mark 111 in the preprocessed image 201 as the center. In this way, the detection mark 111 can be located at the center of the template image 202 (the center of the detection mark 111 in the template image 202 coincides with the image center of the template image 202).

[0090] Furthermore, since the chip 101 is transferred from the original wafer to the substrate 100 , the position of the chip 101 on the substrate 100 may deviate from its theoretical physical position. Figure 7 The actual distribution of some of the chips 101 on the substrate 100 is shown. Figure 8 Shown is the inclusion Figure 7 The first six detection marks 111 in the first row of the substrate image 200 are shown in FIG. Figure 7and Figure 8 As can be seen in FIG. 2 , the chip 101 may be rotated by a certain angle relative to its theoretical physical position on the substrate 100 , resulting in the detection mark 111 in the substrate image 200 being skewed. Fig. 9 For Figure 8 A schematic diagram of extracting the template image 202 from the substrate image 200 in FIG. Fig. 9 As can be seen from the figure, when the chip 101 has a rotation angle deviation relative to its theoretical physical position, the detection mark 111 in the manually captured template image 202 will also be skewed, thereby causing errors in template matching.

[0091] Based on this, before capturing the template image 202 , the template image 202 needs to be corrected first. In this embodiment, any one of the chips 101 on the substrate 100 is selected as the template chip, and then the substrate image 200 corresponding to the template chip is corrected. Fig.10 Schematic diagram of the template chip 101' provided in this embodiment. Fig.10 As shown, the template chip 101' is relative to its theoretical physical position ( Fig.10 The virtual box in the figure) has a rotation angle deviation, and then at least two vertices are arbitrarily selected in the template chip 101' for taking pictures. In this embodiment, the vertex A at the upper right corner of the template chip 101' (the theoretical physical position is a1, the actual physical position is a2) and the vertex B at the lower left corner (the theoretical physical position is b1, the actual physical position is b2) are selected, but it should not be limited to this. In other embodiments, other vertices of the template chip 101' can also be selected.

[0092] Fig.11 The vertex image 301 obtained by shooting the vertex A is as follows: Fig.11 As shown, the center of the field of view of the scanning module is moved to a1 and a photo is taken to obtain the vertex image 301. Fig.11 In the figure, a11 and a21 are the positions of a1 and a2 in the vertex image 301, respectively, that is, a11 is the theoretical pixel position of vertex A in the vertex image 301, a21 is the actual pixel position of vertex A in the vertex image 301, and a21 should be located exactly at the image center of the vertex image 301. Ideally, a1 and a2 should coincide, and a11 and a21 should also coincide, but since the template chip 101' has a rotation angle deviation relative to its theoretical physical position, a1 and a2 do not coincide, and a11 and a21 do not coincide either.

[0093] akin, Fig.12 The vertex image 302 obtained by shooting the vertex B is as follows: Fig.12As shown, the center of the field of view of the scanning module is moved to b1 and a photo is taken to obtain the vertex image 302. Fig.12 , b11 and b21 are the positions of b1 and b2 in the vertex image 302, that is, b11 is the theoretical pixel position of vertex B in the vertex image 302, b21 is the actual pixel position of vertex B in the vertex image 302, and b21 should be located exactly at the image center of the vertex image 302. Ideally, b1 and b2 should coincide, and b11 and b21 should also coincide, but because the template chip 101' has a rotation angle deviation relative to its theoretical physical position, b1 and b2 do not coincide, and b11 and b21 do not coincide either.

[0094] Next, the pixel position deviation between a11 and a21 and the pixel position deviation between b11 and b21 are obtained, and the rotation angle deviation of the template chip 101' relative to its theoretical physical position can be calculated based on this. When the template image 202 is manually intercepted, the substrate image 200 corresponding to the template chip 101' is rotated according to the rotation angle deviation, and the template chip 101' in the rotated substrate image 200 is selected for interception, so as to correct the position of the template chip 101' in the substrate image 200. In this way, it can be ensured that the detection mark 111 in the intercepted template image 202 is not skewed.

[0095] In some embodiments, the template image 202 may also be acquired offline. For example, before scanning the substrate 100 , an image of the detection mark 111 at the same magnification as the lens of the scanning module may be acquired in advance as the template image 202 .

[0096] It should be noted that, since the shapes and sizes of the four detection marks 111 of each chip 101 in this embodiment are the same, it is only necessary to obtain the template image of any one detection mark. When the chip 101 has multiple detection marks (each detection mark has a different shape and / or size), it is necessary to obtain the template image of each detection mark.

[0097] Execute step S400, and obtain the actual physical position of each detection mark 111 according to the actual pixel position and theoretical physical position of each detection mark 111 in the substrate image 200. The actual physical positions of the four detection marks 111 of the chip 101 can represent its actual physical position.

[0098] Specifically, for each detection mark 111, its actual physical position can be calculated in the following manner:

[0099] First, the theoretical pixel position of the detection mark 111 in the substrate image 200 can be converted according to the theoretical physical position of the detection mark 111 and the mapping relationship; then, the pixel position deviation between the theoretical pixel position of the detection mark 111 and its actual pixel position in the substrate image 200 is calculated; then, according to the mapping relationship, the pixel position deviation of the detection mark 111 is converted into a physical position deviation, and the actual physical position of the detection mark 111 can be obtained according to the physical position deviation and the theoretical physical position of the detection mark 111.

[0100] Furthermore, the scanning module usually uses a line scan camera to take pictures. When the scanning speed of the line scan camera does not match the sampling speed frequency, the substrate image 200 may be compressed or stretched, resulting in inaccurate calculated actual physical positions of the detection marks 111. Based on this, after obtaining the actual physical positions of the detection marks 111, it is also necessary to determine whether each of the substrate images 200 is stretched or compressed. When it is determined that the substrate image 200 is stretched or compressed, the actual physical position of the corresponding detection mark 111 needs to be corrected.

[0101] Specifically, according to the actual pixel positions of all the detection marks 111 in the substrate image 200, the actual minimum enclosing rectangle that can frame all the detection marks 111 in the substrate image 200 can be obtained, and according to the theoretical pixel positions of all the detection marks 111 in the substrate image 200, the theoretical minimum enclosing rectangle that can frame all the detection marks 111 in the substrate image 200 can be obtained. When the size of the actual minimum enclosing rectangle is equal to that of the theoretical minimum enclosing rectangle, it can be determined that the substrate image 200 is not stretched or compressed. When the size of the actual minimum enclosing rectangle is not equal to that of the theoretical minimum enclosing rectangle, it can be determined that the substrate image 200 is stretched or compressed.

[0102] Furthermore, the width W1 (dimension along the x direction) and the height H1 (dimension along the y direction) of the actual minimum circumscribed rectangle are obtained, and the width W2 and the height H2 of the theoretical minimum circumscribed rectangle are obtained. When the width W1 of the theoretical minimum circumscribed rectangle is equal to the width W2 of the actual minimum circumscribed rectangle and the height H1 of the theoretical minimum circumscribed rectangle is equal to the height H2 of the actual minimum circumscribed rectangle (W1=W2, H1=H2), it is determined that the substrate image 200 is not compressed or stretched in both the x and y directions; when the width W1 of the theoretical minimum circumscribed rectangle is not equal to the width W2 of the actual minimum circumscribed rectangle (W1≠W2), it is determined that the substrate image 200 is compressed or stretched in the x direction, and the compression ratio or stretching ratio is W1 / W2; when the height H1 of the theoretical minimum circumscribed rectangle is not equal to the height H2 of the actual minimum circumscribed rectangle (H1≠H2), it is determined that the substrate image 200 is compressed or stretched in the y direction, and the compression ratio or stretching ratio is H1 / H2. The substrate image 200 can be compressed or stretched in both the x and y directions.

[0103] Based on this, let the actual physical position of the detection mark 111 in the substrate image 200 be (x i ,y i ), where i is the serial number of the detection mark 111 in the substrate image 200, the x direction is the width direction of the substrate image 200, and y is the height direction of the substrate image 200. When the substrate image 200 is compressed or stretched in the x direction, the x coordinate of the detection mark 111 is corrected using the compression ratio / stretching ratio W1 / W2, and the corrected x coordinate is x i *(W1 / W2); When the substrate image 200 is compressed or stretched in the y direction, the y coordinate of the detection mark 111 is corrected using the compression ratio or stretching ratio H1 / H2, and the corrected y coordinate is y i *(H1 / H2).

[0104] Fig.13 A schematic diagram of three scanning queues is shown in FIG. Fig.13 As shown, the three scanning queues are m1, m2, and m3, and each scanning queue has five substrate images 200. Fig.13 The boxes in the figure are the shooting positions corresponding to each substrate image 200. Ideally, the starting and ending points of the scanning queues m1, m2, and m3 should be consistent, but the positioning error of the motion stage causes the starting and ending points of the scanning queues m1, m2, and m3 to deviate. Fig.13As can be seen, the scan queue m2 has a starting physical position deviation Δy2 relative to the scan queue m1 in the y direction, resulting in the calculated y coordinates of all the detection marks 111 in the scan queue m2 having a starting physical position deviation Δy 2 , the scan queue m3 has an upward starting physical position deviation Δy in the y direction relative to the scan queue m1 3 , resulting in the calculated y coordinates of all the detection marks 111 in the scanning queue m3 having a starting physical position deviation Δy 3 .

[0105] Based on this, in this embodiment, a group of substrate images 200 with corresponding positions are selected from all scanning queues, and the substrate image 200 in the first scanning queue is used as the reference image. The starting physical position deviations of the substrate images 200 in other scanning queues and the reference image 200 are obtained one by one. Then, the actual physical positions of the detection marks 111 in all the substrate images 200 in the scanning queue can be corrected according to the starting physical position deviations of the substrate images 200 in each scanning queue, thereby improving the accuracy of the actual physical position of the detection mark 111. For example, Fig.13 For example, the substrate image 200 in the first row is selected in the scanning queues m1, m2, and m3, and the substrate image 200 in the scanning queue m1 is used as the reference image to obtain the starting physical position deviation (i.e., Δy 2 ), and obtain the starting physical position deviation (ie, Δy 3 ), and then use Δy 2 Correct the actual physical positions of all the detection marks 111 in the five substrate images 200 in the scanning queue m2, using Δy 3 The actual physical positions of all the detection marks 111 in the five substrate images 200 in the scanning queue m3 are corrected.

[0106] Optionally, the actual physical position of the detection mark 111 in the substrate image 200 is assumed to be (x jkn ,y jk ), where j is the sequence number of the scanning queue, k is the sequence number of the substrate image in the scanning queue, and n is the sequence number of the detection mark 111 in the substrate image. j Correct the y coordinates of the detection marks 111 in all the substrate images 200 in the scanning queue, and the corrected y coordinates are y jk+Δy j For example, the actual physical position of the first detection mark 111 in the first row of the substrate image 200 in the scanning queue m2 is (x 211 ,y 21 ), the actual physical position after correction is (x 211 ,y 21 +Δy 2 ).

[0107] Based on this, this embodiment also provides a position detection system. Fig.14 A simplified schematic diagram of the position detection system provided in this embodiment is shown in FIG. Fig.14 As shown, the position detection system comprises:

[0108] The motion module 10 is used to carry a substrate 100, on which a plurality of chips are provided, and each of the chips has at least one detection mark;

[0109] A scanning module 20, configured to scan the entire substrate 100 along a row / column direction to obtain a plurality of substrate images including at least one detection mark;

[0110] an identification module 30, for identifying the detection mark in the substrate image and obtaining an actual pixel position of the detection mark in the substrate image; and

[0111] The position acquisition module 40 is used to acquire the actual physical position of the detection mark according to the actual pixel position and the theoretical physical position of the detection mark, so as to obtain the actual physical position of each chip.

[0112] In summary, in the position detection method and system provided by the embodiment of the present invention, at least one detection mark is set on each chip of the substrate, and then the entire substrate is scanned along the row / column direction to obtain a number of substrate images containing at least one detection mark, identify the detection mark in the substrate image, and obtain the actual pixel position of the detection mark in the substrate image, and obtain the actual physical position of the detection mark according to the actual pixel position and theoretical physical position of the detection mark to obtain the actual physical position of each chip. The present invention can obtain the actual physical position of each chip by setting the detection mark on the chip and obtaining the actual physical position of the detection mark, with high detection efficiency and small error.

[0113] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the various embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0114] It should also 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 embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the scope of protection of the technical solution of the present invention.

[0115] It should also be understood that, unless otherwise specified or indicated, the terms "first", "second", "third", etc. in the specification are merely used to distinguish between the various components, elements, steps, etc. in the specification, and are not used to indicate the logical relationship or sequential relationship between the various components, elements, steps, etc.

[0116] It should also be recognized that the terms described herein are only used to describe specific embodiments and are not intended to limit the scope of the invention. It should be noted that the singular forms "a" and "an" used herein and in the appended claims include plural references unless the context clearly indicates otherwise. For example, a reference to "a step" or "a device" means a reference to one or more steps or devices, and may include secondary steps and secondary devices. All conjunctions used should be understood in the broadest sense. And, the word "or" should be understood to have the definition of a logical "or", rather than a logical "exclusive or", unless the context clearly indicates otherwise. In addition, the implementation of the method and / or device in the embodiments of the present invention may include performing the selected task manually, automatically, or in combination.

Claims

1. A position detection method, characterized in that: include: Providing a substrate, wherein the substrate has a plurality of chips, and each of the chips has at least one detection mark; Scanning the entire substrate along a row / column direction to obtain a plurality of substrate images including at least one detection mark; Identify the detection mark in the substrate image, and obtain the actual pixel position of the detection mark in the substrate image; as well as, The actual physical position of the detection mark is acquired according to the actual pixel position and the theoretical physical position of the detection mark to obtain the actual physical position of each chip.

2. The position detection method according to claim 1, characterized in that: The detection mark is identified in the substrate image by using a template matching algorithm using the template image of the detection mark; and / or the detection mark is directly outlined in the substrate image.

3. The position detection method according to claim 2, characterized in that: The template image is automatically extracted from any of the substrate images using an automatic mark extraction algorithm.

4. The position detection method according to claim 3, characterized in that: The automatic label extraction algorithm includes at least one of an extraction algorithm based on image saliency, an extraction algorithm based on information entropy, and an extraction algorithm based on matrix principal components.

5. The position detection method according to claim 2, characterized in that: The template image is cut out from any of the substrate images.

6. The position detection method according to claim 5, characterized in that: The step of intercepting the template image comprises: Extracting a pre-processed image including the detection mark from any of the substrate images; and The detection mark is identified in the pre-processed image, and the template image including the detection mark is cut out from the pre-processed image with the center of the identified detection mark as the center.

7. The position detection method according to claim 5 or 6, characterized in that: Before intercepting the template image, the position detection method further includes: Selecting one of the chips on the substrate as a template chip, moving the field of view center of the scanning module to the theoretical physical positions of at least two vertices of the template chip one by one and taking photos to obtain at least two vertex images; Obtaining the rotation angle deviation of the template chip relative to its theoretical physical position according to the pixel position deviation between the vertex in each vertex image and the image center; and The substrate image corresponding to the template chip is rotated according to the rotation angle deviation, and the template chip in the rotated substrate image is selected for interception.

8. The position detection method according to claim 2, characterized in that: Before scanning the substrate, an image of the detection mark at the same magnification as that of a lens of a scanning module is acquired in advance as the template image.

9. The position detection method according to claim 2, characterized in that: The template matching algorithm includes the ncc algorithm.

10. The position detection method according to claim 1, characterized in that: Before scanning the substrate, a mapping relationship between a coordinate system of the substrate and a coordinate system of a scanning module is acquired.

11. The position detection method according to claim 10, characterized in that: The step of obtaining the actual physical position of the detection mark according to the actual pixel position and the theoretical physical position of the detection mark comprises: Acquiring a theoretical pixel position of the detection mark according to the theoretical physical position of the detection mark and the mapping relationship; Obtaining a pixel position deviation between a theoretical pixel position and an actual pixel position of the detection mark; converting the pixel position deviation of the detection mark into a physical position deviation according to the mapping relationship; and The actual physical position of the detection mark is obtained according to the physical position deviation and the theoretical physical position of the detection mark.

12. The position detection method according to claim 10 or 11, characterized in that: After obtaining the actual physical position of the detection mark, the position detection method further includes: Acquiring a theoretical pixel position of the detection mark according to the theoretical physical position of the detection mark and the mapping relationship; Determining whether the substrate image is stretched or compressed according to the theoretical pixel positions and actual pixel positions of all the detection marks in the substrate image; and The actual physical position of the detection mark in the substrate image is corrected according to the stretching ratio or the compression ratio of the substrate image.

13. The position detection method according to claim 12, characterized in that: The step of determining whether the substrate image is stretched or compressed comprises: Acquire an actual minimum circumscribed rectangle capable of enclosing all the detection marks in the substrate image according to actual pixel positions of all the detection marks in the substrate image; Acquire a theoretical minimum circumscribed rectangle capable of enclosing all the detection marks in the substrate image according to theoretical pixel positions of all the detection marks in the substrate image; and When the actual minimum circumscribed rectangle is equal to the theoretical minimum circumscribed rectangle in size, it is determined that the substrate image is not stretched or compressed; when the actual minimum circumscribed rectangle is not equal to the theoretical minimum circumscribed rectangle in size, it is determined that the substrate image is stretched or compressed.

14. The position detection method according to claim 13, characterized in that: When the width W1 of the theoretical minimum circumscribed rectangle is not equal to the width W2 of the actual minimum circumscribed rectangle, it is determined that the substrate image is stretched or compressed in the width direction, and the stretching ratio or compression ratio is W1 / W2; when the height H1 of the theoretical minimum circumscribed rectangle is not equal to the height H2 of the actual minimum circumscribed rectangle, it is determined that the substrate image is stretched or compressed in the height direction, and the stretching ratio or compression ratio is H1 / H2.

15. The position detection method according to claim 14, characterized in that The actual physical position of the detection mark in the substrate image is (x i ,y i ), wherein i is the serial number of the detection mark in the substrate image, the x direction is the width direction of the substrate image, and the y direction is the height direction of the substrate image; and, When the substrate image is stretched or compressed in the width direction, the x coordinate of the detection mark i is corrected using the stretching ratio or compression ratio W1 / W2 of the substrate image. The corrected x coordinate is x i *(W1 / W2), when the substrate image is stretched or compressed in the height direction, the y coordinate of the detection mark i is corrected by the stretching ratio or compression ratio H1 / H2 of the substrate image, and the corrected y coordinate is y i *(H1 / H2).

16. The position detection method according to claim 1, characterized in that: After obtaining the actual physical position of the detection mark, the position detection method further includes: A group of substrate images with corresponding positions are selected from all scanning queues, and the substrate image in the first scanning queue is used as a reference image, and the starting physical position deviations of the substrate images in other scanning queues and the reference image are obtained one by one; and The actual physical positions of the detection marks in all the substrate images in each scanning queue are corrected according to the starting physical position deviation of the substrate image in the scanning queue.

17. The position detection method according to claim 16, characterized in that The actual physical position of the detection mark in the substrate image is (x jkn ,y jk ), wherein j is the sequence number of the scanning queue, k is the sequence number of the substrate image in the scanning queue, n is the sequence number of the detection mark in the substrate image, and y is the scanning direction; According to the starting physical position deviation Δy of the substrate image of each scanning queue j Correct the y coordinates of the detection marks in all the substrate images in the scanning queue, and the corrected y coordinates are y jk +Δy j .

18. A position detection system, characterized in that: include: A motion module, used for carrying a substrate, the substrate having a plurality of chips, each of the chips having at least one detection mark; A scanning module, used for scanning the entire substrate along a row / column direction to obtain a plurality of substrate images including at least one detection mark; an identification module, used for identifying the detection mark in the substrate image and obtaining the actual pixel position of the detection mark in the substrate image; as well as, The position acquisition module is used to acquire the actual physical position of the detection mark according to the actual pixel position and the theoretical physical position of the detection mark, so as to obtain the actual physical position of each chip.

Images