JP2009162718A - Substrate inspection apparatus and method for setting inspection area - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To set an inspection area on an image of a substrate without the need for a microscope even when the image of the substrate does not show any feature of the substrate as a detailed appearance. <P>SOLUTION: An image processing section 22 sets an inspection area on an image of a substrate in accordance with inspection image information pp based on two-dimensional image kk outputted from an image-taking section 51 and inspection manufacturing step/type information mm including design information defining the inspection area and outputted from the manufacturing step/type information acquiring section 12. In the process, the image processing section 22 generates a design image indicative of the substrate in accordance with the design information and cross-checks the design image with the image of the substrate as indicated by the inspection image information pp. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an apparatus and method for inspecting a substrate.

  In the manufacturing process of a substrate such as a flat panel display (FPD) such as an LCD (Liquid Crystal Display) and a PDP (Plasma Display Panel) and a semiconductor wafer, an inspection is performed on the substrate. As a result of the inspection, whether the substrate is good or bad is determined according to the presence or absence of a defect or the type of defect found.

  In the substrate manufacturing process, a plurality of types of inspection may be performed depending on the type of defect to be detected. For example, an inspection called “micro inspection” for detecting a minute defect using a microscope and an inspection called “macro inspection” for detecting a defect having a relatively large area are used in combination. The detection objects of the micro inspection and the macro inspection are sometimes called “micro defects” and “macro defects”, respectively.

  In general, a macro defect is a defect that has a size that cannot be found by microscopic observation using a microscope on the substrate to be inspected, and a single large defect that is caused by the concentration of similar micro defects locally. It refers to defects that are observed. Macro defects are generally regarded as defects having a relatively large area.

  For example, a relatively large foreign matter adhering to the substrate during the manufacturing process and a relatively large scratch generated on the substrate are examples of macro defects. Also, exposure defects over a certain size range are examples of macro defects. Further, in the process of forming a film such as a photoresist on the substrate, unevenness due to the change in film thickness is also a detection target of the macro inspection. For example, in order to detect unevenness associated with a change in film thickness, it is effective to observe the entire substrate rather than local observation such as micro inspection.

  Since macro defects are relatively large, in conventional macro inspection, the operator (operator) discovers macro defects visually while moving and / or rotating the substrate to be inspected or the illumination system that illuminates the substrate. Was.

  In addition, a macro inspection apparatus that automatically performs a macro inspection has been developed. The macro inspection apparatus captures a macro defect that has been observed visually with an imaging sensor. For example, the macro inspection apparatus generates an image of the entire substrate by imaging the entire substrate with a single imaging, and detects a macro defect based on the image. Some macro inspection apparatuses have a function of determining whether a substrate is acceptable according to the presence or absence of a macro defect and controlling whether the substrate is advanced to the next process.

  In this way, various inspections such as macro inspection are performed in the substrate manufacturing process. One of the important elements for realizing accurate inspection is the area to be inspected (hereinafter referred to as “inspection area”). There is a problem of how to set. Currently, when an operator sets information indicating a recipe (inspection condition), a method of manually specifying an inspection region by using a pointing device or the like while viewing an image obtained by imaging a substrate to be inspected. It is common. However, such manual operation has the following problems.

  For example, in an FPD manufacturing process, a plurality of display panels are formed on a single glass substrate and then cut into individual display panels. Similarly, one semiconductor wafer is divided into a plurality of chips along a scribe line. With respect to certain types of defects, the criteria for determining whether or not a substrate is good may differ depending on whether a region where a circuit is formed or a region other than the region is defective.

  Therefore, in order to perform inspection according to the attribute of the region on the substrate, the operator sets, for example, the region where the circuit is formed as the first inspection region, and the first recipe corresponding to the first inspection region. Set. In addition, the operator sets an area in which no circuit is formed as the second inspection area, and sets a second recipe corresponding to the second inspection area. The operator may set an inspection area and a recipe, respectively, according to three or more types of attributes.

  In addition, as the density of circuits formed on a substrate increases, that is, the tendency to miniaturization increases, the required inspection resolution of an object to be inspected has also increased. That is, when a plurality of inspection areas and recipes are to be set according to the attributes of the area on the substrate, it has been required to set the boundaries between the plurality of inspection areas more accurately. In other words, the accuracy required for manual operation for setting the inspection area is increasing.

  However, manual operation with high accuracy requires a long time or requires skill of the operator. Therefore, there is a problem that the time required for setting the inspection region and the accuracy of setting the inspection region, and hence the time required for the macro inspection and the accuracy of the macro inspection vary depending on the operator.

  In order to solve the above problem, it is desired to automate the setting of the inspection area in the macro inspection apparatus. For this purpose, the macro inspection apparatus needs to recognize an area on the image corresponding to a specific area on the substrate. That is, the macro inspection apparatus needs to recognize the correspondence between the position represented in the board design information and the position on the image obtained by imaging the board. Hereinafter, the process of recognizing the correspondence between the position represented in the board design information and the position on the image obtained by imaging the board is referred to as “positioning process”.

  For example, each chip region formed on the semiconductor wafer includes an alignment mark used for alignment in a stepper. The position of the alignment mark is predetermined by design information, and when the substrate is imaged using a microscope, the alignment mark appears in the captured image. Therefore, it is possible to perform alignment processing using the alignment mark. Not only the alignment mark but also the shape of the circuit pattern, for example, can be used for the alignment process.

  For example, a technique is disclosed in which design data is taken into a recipe extraction system, the recipe extraction system recognizes an instance of a target structure such as an alignment site, and configures recipe parameters according to this (for example, Patent Document 1). reference.). Furthermore, threshold processing based on regions is also disclosed (for example, see Patent Document 1).

In addition, a method for creating an inspection recipe for inspecting a plurality of inspection areas having different image brightness due to pattern density in a circuit pattern inspection of a semiconductor wafer is disclosed (for example, see Patent Document 2). In this method, an inspection threshold is registered for each inspection region. In addition, alignment based on the alignment mark is performed using an SEM (Scanning Electron Microscope) image obtained by imaging the semiconductor wafer.
JP 2005-514774 A JP 2007-67170 A

  However, a minute shape such as an alignment mark or a circuit pattern does not always appear visually as an appearance in an image obtained by imaging the substrate. For example, when a substrate is imaged by a macro inspection apparatus, the resolution of an image is generally lower than that of a microscope, so that a minute shape may not appear visually as an appearance.

  In this case, even if the design information includes information on the position and shape of the alignment mark or circuit pattern, it has not been possible to perform the alignment process using the alignment mark or circuit pattern as a reference. In other words, conventionally, it has been impossible to specify an area to be inspected on an image obtained by imaging a substrate based only on an image obtained by imaging the substrate and design information.

  In order to solve this problem, in order to enable alignment processing using a minute shape, a microscope for alignment processing is attached to the substrate inspection apparatus, even if it is not necessary for inspection itself, and high A method of imaging the substrate with resolution is considered as an option. According to this method, it is possible to perform alignment processing with high accuracy based on a minute shape captured in an image with high resolution. However, this method requires a device such as a microscope that is unnecessary for the inspection itself, and the positional relationship between the observation system equipped with the microscope for the alignment process and the observation system for the inspection is precise. It is not preferable in that it is necessary to set or measure.

  Therefore, the present invention captures a substrate without requiring a device for performing imaging with high resolution even if not all of the features of the substrate are visually shown as a detailed appearance in the image of the substrate. It is an object of the present invention to provide an apparatus and method capable of setting an area to be inspected on a processed image.

  According to one aspect of the present invention, there is provided a substrate inspection apparatus that sets a region on a substrate image obtained by imaging a substrate to be inspected as a substrate image inspection region to be inspected. According to another aspect of the present invention, there is provided an inspection area setting method executed by a computer that functions in the same manner as the substrate inspection apparatus.

  The substrate inspection apparatus includes an image acquisition unit that acquires the substrate image, a design information acquisition unit that acquires design information of the substrate, including a definition of a substrate inspection region that is an inspection target on the substrate, and the design information. The design image representing the substrate is acquired from the image, the design image and the substrate image are collated, and the substrate image corresponding to the substrate inspection region based on the collation result and the design information Inspection region setting means for setting the upper region as the substrate image inspection region.

  The inspection area setting means is a design image generation means for generating the design image having the same resolution as the board image based on the design information, and a difference indicating a relative positional shift and rotation between the design image and the board image. You may provide the alignment processing means which calculates a value. Then, the inspection area setting means is arranged on the board image corresponding to the board inspection area based on the area on the design image defined by the design information corresponding to the board inspection area and the difference value. May be determined, and the determined area may be set as the substrate image inspection area.

  According to the present invention, an area on a board image corresponding to a board inspection area can be set as a board image inspection area without requiring manual operation.

  According to the present invention, the inspection area setting unit collates the design image with the substrate image without assuming that a specific minute pattern such as an alignment mark appears in the substrate image. Therefore, even if not all the features of the board are visually shown as a detailed appearance in the board image, the inspection area setting means collates the design image with the board image using the feature that appears in the board image, and the board Set the image inspection area.

  Therefore, according to the present invention, a device such as a microscope is not required for setting the substrate image inspection area.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the following, an embodiment in which the substrate to be inspected is a semiconductor wafer will be described, but the object to be inspected may be an FPD substrate or the like.

FIG. 1 is a functional configuration block diagram of a substrate inspection apparatus according to an embodiment. The substrate inspection apparatus 10 in FIG. 1 is an example of a macro inspection apparatus.
The substrate inspection apparatus 10 includes an apparatus control unit 11 that controls the entire substrate inspection apparatus 10, an imaging unit 51 that performs an imaging process for acquiring an image of a semiconductor wafer that is a substrate to be inspected, And an inspection processing unit 52 for setting an inspection area to be inspected and performing the inspection.

  Further, outside the substrate inspection apparatus 10, an input unit 54 for accepting an external input given as an instruction to the substrate inspection apparatus 10, details of processing in the substrate inspection apparatus 10, the state of the substrate inspection apparatus 10, etc. Is displayed or displayed. The device control unit 11 is connected to the display unit 53 and the input unit 54.

  The input unit 54 is realized by, for example, a keyboard, a pointing device such as a mouse, a microphone, or a combination thereof. The display unit 53 may be any output device such as a display, a speaker, a printer, and a communication interface for transmitting data to an external device. Therefore, although the display unit 53 is also expressed as an “output unit” in FIG. 1, the display unit 53 of the present embodiment is a display.

Hereinafter, an image of a semiconductor wafer imaged by the imaging unit 51 is referred to as a “macro image”. A macro image is an example of a substrate image obtained by imaging a substrate to be inspected.
The imaging unit 51 includes an inspection stage 18, an illumination unit 15, an imaging sensor 16, an illumination system control unit 13, an imaging sensor control unit 14, a stage control unit 17, and an image capture unit 19. The imaging unit 51 is an example of an image acquisition unit that acquires a substrate image.

  That is, the imaging unit 51 that acquires a macro image includes only a macro observation system, and does not include a micro observation system including a microscope for magnifying the substrate. However, even if a micro observation system including a microscope is not provided, the substrate inspection apparatus 10 can determine an inspection region and perform a macro inspection as follows.

The inspection stage 18 is a stage on which a substrate, that is, a semiconductor wafer is placed, and has a drive unit such as a motor. The inspection stage 18 is controlled by the stage controller 17.
The illumination unit 15 applies illumination light to the substrate in a state where the illumination unit 15 is inclined with respect to the substrate placed on the inspection stage 18. The illumination unit 15 is controlled by the illumination system control unit 13.

  The imaging sensor 16 is a one-dimensional line sensor including photoelectric conversion elements arranged in a line shape, and receives the light irradiated from the illumination unit 15 and reflected by the surface of the substrate to generate line image information jj, thereby generating an image. Output to the capture unit 19.

  The imaging sensor 16 may be, for example, a CCD (Charge Coupled Device) image sensor or a CMOS (Complementary Mental-Oxide Semiconductor) image sensor. The image sensor 16 of the present embodiment is a sensor that captures a monochrome luminance image, but a color image sensor may be used in another embodiment.

The image sensor 16 is controlled by the image sensor controller 14.
When the stage control unit 17 controls and moves the inspection stage 18, or when the imaging sensor control unit 14 controls and moves the imaging sensor 16, imaging includes the substrate transport system including the inspection stage 18 and the imaging sensor 16. The relative movement of the system takes place.

  Here, if the direction of the coordinate axes of the XYZ coordinate system is appropriately determined, the surface of the stage control unit 17 is parallel to the XY plane, and the direction in which the relative movement is performed is the Y direction. The image sensor 16 is installed in an appropriate direction so that the surface of the substrate is imaged in a line in the X direction corresponding to the photoelectric conversion elements arranged in a line.

  The imaging unit 51 repeats the imaging by the imaging sensor 16 and the acquisition of the line image information jj by the image capturing unit 19 while relatively moving the base conveyance system and the imaging system in the Y direction. Since the line image information jj is data of an image obtained by imaging one line in the X direction, the image capturing unit 19 represents the surface of the substrate 2 by sequentially arranging and combining sequentially acquired images of a plurality of lines in the Y direction. The macro image which is a dimensional image can be generated.

  Hereinafter, the information representing the macro image is described as “two-dimensional image information kk”. The two-dimensional image information kk is used in various macro inspection processes such as defect extraction, defect classification, and chip / substrate quality determination. For example, in defect classification, defects are classified based on the area, shape, position, and the like of the extracted defects. In addition, the determination of “defective” in the substrate quality determination indicates that the substrate should not be advanced to the next step.

  Note that the X and Y directions in the XYZ coordinate system representing the position of the apparatus such as the stage control unit 17 correspond to the X and Y directions in the two-dimensional image information kk, respectively. Hereinafter, since there is no fear of confusion, they are expressed as “X direction” and “Y direction” without distinction.

  The image capturing unit 19 outputs the generated two-dimensional image information kk to the inspection processing unit 52. Then, the inspection processing unit 52 sets an inspection region to be inspected on the macro image based on the two-dimensional image information kk, and executes the macro inspection according to the setting. The inspection processing unit 52 outputs inspection result information ss as inspection area information. The inspection processing unit 52 also performs various processes related to the macro inspection after setting the inspection area, for example, processes such as defect extraction and classification, and substrate pass / fail determination.

  The inspection processing unit 52 includes a manufacturing process / product type information acquisition unit 12, an image storage unit 20, an image processing unit 22, and an inspection information generation unit 23. The image storage unit 20 may be a volatile or nonvolatile semiconductor memory or a hard disk device. The manufacturing process / product information acquisition unit 12 is an example of the design information acquisition unit.

  In the substrate inspection apparatus 10, each unit other than the illumination unit 15, the imaging sensor 16, and the inspection stage 18 may be realized by a dedicated hardware circuit, for example, by a computer and a program having a general configuration. May be.

  For example, a CPU (Central Processing Unit), a nonvolatile memory such as a ROM (Read Only Memory), a RAM (Random Access Memory) used as a working area, an external storage device such as a hard disk device, various external devices and a network Each unit other than the illumination unit 15, the imaging sensor 16, and the inspection stage 18 may be realized by a computer that includes a connection interface with each other and connected to each other via a bus.

  That is, the function of each unit may be realized by the CPU executing a program stored in a ROM, a computer-readable portable storage medium, or an external storage device. Further, the RAM or hard disk device provided in the computer may realize the image storage unit 20. The connection interface includes, for example, an interface between various devices such as the illumination unit 15, the image sensor 16, the inspection stage 18, the display unit 53, and the input unit 54 and a PC, and a LAN to which a server (not shown) is connected. Includes interfaces with networks such as (Local Area Network).

  More specifically, for example, the image capturing unit 19 may be realized by an image capture board as a connection interface with the imaging sensor 16. Further, the input unit 54 and the display unit 53 may be a keyboard or a display of another computer connected via a network.

The configuration of the substrate inspection apparatus 10 is as described above. Next, the operation of the substrate inspection apparatus 10 will be described.
A system management unit (not shown) directly connected to the substrate inspection apparatus 10 or connected via a network outputs control information aa for performing inspection to the substrate inspection apparatus 10. The system management unit manages the substrate inspection apparatus 10 and other apparatuses in order to manage the manufacturing process.

  The control information aa is received by the apparatus control unit 11 that performs internal control of the board inspection apparatus 10. The control information aa includes information indicating that a substrate to be inspected is loaded into the substrate inspection apparatus 10.

  Upon receiving the control information aa, the apparatus control unit 11 instructs the illumination system control unit 13, the imaging sensor control unit 14, and the stage control unit 17 to make preparations for loading the substrate. Information is sent bi-directionally between these three control units and the device control unit 11.

  Hereinafter, information exchanged between the device control unit 11 and the illumination system control unit 13 is referred to as “illumination control information cc”, and information exchanged between the device control unit 11 and the imaging sensor control unit 14 is referred to as “imaging sensor control information dd”. The information exchanged between the apparatus control unit 11 and the stage control unit 17 is represented as “stage control information ee”.

The illumination control information cc, the imaging sensor control information dd, and the stage control information ee at this time each include an instruction for loading the substrate.
The illumination system control unit 13, the image sensor control unit 14, and the stage control unit 17 prepare for imaging a substrate for inspection based on an instruction from the apparatus control unit 11. When the preparation is completed, the three control units respectively notify the apparatus control unit 11 of information indicating that “inspection preparation is complete” as illumination control information cc, imaging sensor control information dd, or stage control information ee.

  Upon receiving notification of inspection preparation completion from all three control units, the apparatus control unit 11 recognizes that inspection preparation has been completed. For example, control for moving the inspection stage 18 or the image sensor 16 may be performed as preparation for inspection so that the relative positions of the substrate transport system and the image capturing system are in a predetermined initial state.

  When the apparatus control unit 11 recognizes that the inspection preparation is completed, the device control unit 11 instructs the illumination system control unit 13, the image sensor control unit 14, and the stage control unit 17 to perform an imaging process for acquiring a macro image, respectively. To do. This instruction is also output from the apparatus control unit 11 as a kind of illumination control information cc, imaging sensor control information dd, and stage control information ee.

  Upon receiving this instruction, the illumination system control unit 13 controls the illumination unit 15, the imaging sensor control unit 14 controls the imaging sensor 16, and the stage control unit 17 controls the inspection stage 18. In the present embodiment, for the sake of simplicity, the following description will be made assuming that relative movement between the substrate transport system and the imaging system is performed by moving the inspection stage 18 in the Y direction. Specifically, the imaging process is as follows.

  The illumination system control unit 13 generates illumination information ff for designating the illumination light amount and outputs the illumination information ff to the illumination unit 15 to control the illumination unit 15. The illumination unit 15 irradiates the substrate placed on the inspection stage 18 with illumination light according to the control based on the illumination information ff.

  Information necessary for the illumination system control unit 13 to determine the illumination information ff may be included in the illumination control information cc as an execution instruction of the imaging process from the device control unit 11. When generating the illumination information ff, the illumination system control unit 13 may use information that is determined in advance and stored in a hard disk device or the like.

  When performing inspection using a diffraction image, if a certain type of macro defect exists, the first-order diffracted light may be stronger than the zero-order diffracted light. Is valid. Therefore, the illumination unit 15 irradiates the substrate with illumination light from an oblique direction different from the Z direction. The illumination information ff may include information for designating an angle at which the illumination light is irradiated when the illumination unit 15 is movable. When the illumination unit 15 includes a plurality of light sources having different wavelengths, the illumination information wavelength ff Information to be specified may be included.

  The stage control unit 17 also generates stage drive information hh for driving the inspection stage 18 on which the substrate to be inspected is mounted in the Y direction, and outputs the stage drive information hh to the driving unit of the inspection stage 18. Control movement. That is, the stage control unit 17 moves the inspection stage 18 in the Y direction perpendicular to the X direction that is the direction of the one-dimensional line imaged by the imaging sensor 16, thereby moving the substrate transport system and the imaging system relative to each other. To realize.

The stage drive information hh may include, for example, an initial speed, acceleration, and arrival speed at the time of driving, and may include a voltage value for controlling the driving speed.
Information necessary for the stage control unit 17 to determine the stage drive information hh may be included in the stage control information ee as an execution instruction of the imaging process from the apparatus control unit 11. When determining the stage drive information hh, the stage control unit 17 may use information that is determined in advance and stored in a hard disk device or the like.

  Further, the imaging sensor control unit 14 generates imaging information gg including an exposure time at the time of imaging, a gain, an imaging start timing, a line transfer timing, and the like, and outputs the imaging information gg to the imaging sensor 16 to control the imaging sensor 16. . The imaging sensor 16 instructed by the imaging sensor control unit 14 to repeat imaging in accordance with the imaging information gg. The imaging sensor 16 generates line image information jj for one line in the X direction every time imaging is performed, and outputs the line image information jj to the image capturing unit 19.

  Information necessary for the imaging sensor control unit 14 to determine the imaging information gg may be included in the imaging sensor control information dd as an instruction to execute the imaging process from the apparatus control unit 11. The image sensor control unit 14 may use information predetermined and stored in a hard disk device or the like when generating the image information gg.

  Note that in order to generate an appropriate macro image from the line image information jj, the speed in the Y direction of the inspection stage 18, the imaging interval of the imaging sensor 16, and the resolution of the line image information jj must match each other. is there. For example, when one pixel in the line image information jj corresponds to the range of p [mm] × p [mm] of the substrate and the imaging interval specified by the imaging information gg is d [seconds], the stage drive information hh The speed in the Y direction of the inspection stage 18 should be specified as p / d [mm / second].

  The apparatus control unit 11 specifies the image sensor control information dd and the stage control information ee so that such consistency is maintained. As a result, consistency between the imaging information gg output from the imaging sensor control unit 14, the stage drive information hh output from the stage control unit 17, and the resolution of the line image information jj is also maintained.

  As described above, every time imaging for each line is repeated, the line image information jj is sequentially acquired by the imaging sensor 16 and is captured by the image capturing unit 19. The image capturing unit 19 combines the line image information jj corresponding to a plurality of lines to generate a macro image that is a two-dimensional image related to the substrate to be inspected.

  The image capturing unit 19 not only simply arranges images of a plurality of lines in order based on the line image information jj, but also shading correction for correcting brightness, distortion correction for correcting geometric distortion, and the like. The image processing related to the correction of the imaging system is also performed. The two-dimensional image after these corrections are performed is a macro image, and the data of the macro image is included in the two-dimensional image information kk.

FIG. 2 is an example of a macro image generated by the image capturing unit 19.
A macro image 100 in FIG. 2 is an image generated by the image capturing unit 19 based on the line image information jj obtained by imaging the semiconductor wafer 101 placed on the inspection stage 18 by the imaging sensor 16. The two-dimensional image information kk representing the macro image 100 is used in the processing in the inspection processing unit 52. In the macro image 100, chips 102 periodically arranged on the semiconductor wafer 101 are shown, and a V-shaped notch 109a that serves as a reference for the orientation of the semiconductor wafer 101 is also shown.

  The resolution of the macro image 100 is so low that the minute circuit pattern and alignment mark inside each chip 102 cannot be recognized by the image processing unit 22 and the inspection information generation unit 23. For example, the inspection processing unit 52 can recognize from the macro image 100 in FIG. 2 by image processing that each chip 102 includes three regions having different luminances according to the density of the circuit pattern. However, the resolution of the macro image 100 is low enough that the inspection processing unit 52 cannot recognize fine wiring patterns in the three regions by image processing. Therefore, when using the macro image 100 acquired for the macro inspection, the inspection processing unit 52 cannot perform alignment processing using a minute shape such as an alignment mark.

Returning to the description of FIG. 1, the processing in the inspection processing unit 52 using the two-dimensional image information kk will be described.
In the present embodiment, in parallel with the acquisition of the two-dimensional image information kk by imaging in the imaging unit 51 as described above, the substrate inspection apparatus 10 is necessary for the inspection in the manufacturing process / product information acquisition unit 12. Information is received from a system management unit (not shown). Here, “information necessary for inspection” is specifically inspection object information bb, which substrate of the current inspection object substrate has been processed, and what kind of product It contains information such as the substrate. The board inspection apparatus 10 can recognize the items to be inspected and the inspection conditions based on the inspection object information bb as follows.

  The manufacturing process / product information acquisition unit 12 converts the inspection object information bb into information in a format that can be used in defect extraction processing to be performed later (hereinafter referred to as “inspection manufacturing process / product information mm”). To do. For example, the manufacturing process / product information acquisition unit 12 converts information about the product included in the inspection object information bb into design information of the semiconductor wafer to be inspected.

  The design information of the semiconductor wafer will be described in detail later with reference to FIG. 3. For example, the size, the interval, the vertical arrangement position, and the horizontal arrangement position regarding both the chip of the semiconductor wafer and the shot at the time of exposure, Information. Items to be inspected and inspection conditions depend on design information of the semiconductor wafer.

  For example, when the design information is embedded in the inspection object information bb, the conversion from the inspection object information bb to the inspection manufacturing process / product information mm by the manufacturing process / product information acquisition unit 12 is performed from the inspection object information bb. It may be a process of extracting design information, a process of acquiring a design information by searching a database (not shown) using a code of a product type included in the inspection object information bb as a search key, or other processes. The manufacturing process / product information acquisition unit 12 outputs the manufacturing process / product information mm for inspection acquired by the conversion to each unit of the device control unit 11, the image storage unit 20, the image processing unit 22, and the inspection information generation unit 23. .

  Further, the two-dimensional image information kk representing the two-dimensional image, that is, the macro image 100 acquired by the image capturing unit 19 is output to the image storage unit 20 in the inspection processing unit 52. The image storage unit 20 stores the two-dimensional image information kk and the inspection manufacturing process / product information mm in association with each other.

  For example, the system management unit (not shown) controls the output timing of the control information aa and the inspection object information bb, whereby the two-dimensional image information kk and the inspection manufacturing process / product type information are stored in the image storage unit 20 at a predetermined timing. The substrate inspection apparatus 10 may be indirectly controlled so that mm is input. Then, the image storage unit 20 may associate the two-dimensional image information kk and the inspection manufacturing process / product information mm based on the input timing.

  Alternatively, the system management unit may embed a common identification number or the like in the control information aa and the inspection target information bb, and the apparatus control unit 11 may notify the identification number to each unit of the imaging unit 51 and the inspection processing unit 52. Accordingly, the image capturing unit 19 embeds the identification number in the two-dimensional image information kk, the manufacturing process / product information acquisition unit 12 embeds the identification number in the inspection manufacturing process / product information mm, and the image storage unit 20 uses the identification number. The corresponding two-dimensional image information kk can be associated with the inspection manufacturing process / product information mm.

  When the image storage unit 20 completes the process of storing the two-dimensional image information kk and the inspection manufacturing process / product type information mm in association with each other, the image storage unit 20 outputs the image storage control information nn to the apparatus control unit 11. The image storage control information nn is information exchanged bidirectionally between the apparatus control unit 11 and the image storage unit 20 with respect to operation timing control. The contents of the image storage control information nn vary depending on the progress of the process, but here is “notification of image information acquisition”, that is, a notification that the acquisition of the two-dimensional image information kk is completed.

  Upon receiving the image storage control information nn indicating “image information acquisition completion”, the apparatus control unit 11 sends the illumination control information cc and the image sensor to the illumination system control unit 13, the image sensor control unit 14, and the stage control unit 17, respectively. As the control information dd and the stage control information ee, control information for notifying “image information acquisition completion” is output. Thereby, the apparatus control unit 11 causes each unit in the imaging unit 51 to complete an operation for image acquisition.

  In addition, when the image capturing unit 51 completes the operation for acquiring an image, the apparatus control unit 11 recognizes that an inspection based on the acquired two-dimensional image information kk should be started. Therefore, the apparatus control unit 11 outputs, to the image storage unit 20, “image inspection start”, that is, notification that the inspection using the two-dimensional image information kk should be started as the image storage control information nn.

  When the image storage unit 20 receives the image storage control information nn for notifying “start image inspection”, the image storage unit 20 performs inspection based on the two-dimensional image information kk and the manufacturing process / product information mm for inspection stored in association with each other. The image information pp is generated, and the inspection image information pp is output to the image processing unit 22 and the inspection information generation unit 23.

  The inspection image information pp is information necessary for determining the inspection conditions and the like, information on which process in the manufacturing process is completed, and information on the inspection target board. Information on the product type, data on the macro image 100, and the like are included.

  As described above, the resolution of the macro image 100 is so low that a minute pattern such as an alignment mark cannot be recognized by image processing. Therefore, in this embodiment, the image processing unit 22 and the inspection information generation unit 23 determine a region to be inspected in the macro image 100 without using a minute pattern such as an alignment mark. Details of the image processing unit 22 and the inspection information generation unit 23 will be described later, but the outline is as follows.

  The image processing unit 22 performs collation processing based on the inspection manufacturing process / product information mm and the inspection image information pp, and outputs image processing result information rr obtained as a result of the collation to the inspection information generation unit 23. The image processing result information rr is, for example, translation conversion and rotation conversion (hereinafter referred to as “translation / rotation conversion”) as information indicating a positional shift and rotation between the design information represented by the inspection manufacturing process / product information mm and the actual macro image 100. Data of the macro image 100 whose position and direction are corrected by “rotation conversion”).

  The inspection information generation unit 23 performs a process of setting an inspection region in order to perform inspection under different conditions for each region corresponding to an attribute such as a chip or a scribe line on a semiconductor wafer. Therefore, the inspection information generation unit 23 refers to the inspection manufacturing process / product type mm and the image processing result information rr and associates the inspection sensitivity corresponding to the attribute on the translated / rotated macro image 100. A plurality of inspection areas are set.

  The inspection information generation unit 23 generates inspection result information ss including attributes, inspection sensitivity, and inspection area information associated with each other as a setting result, and outputs the inspection result information ss to a system management unit (not shown). The system management unit creates a macro inspection recipe based on the inspection result information ss, and outputs inspection object information bb including recipe information to the manufacturing process / product information acquisition unit 12.

  In the present embodiment, the inspection information generation unit 23 executes the macro inspection according to the recipe using the data of the macro image 100 subjected to translation / rotation conversion included in the image processing result information rr, for example. Upon receiving the inspection object information bb including recipe information, the manufacturing process / product information acquisition unit 12 instructs the inspection information generation unit 23 to execute the macro inspection, and the inspection information generation unit 23 displays the result of the macro inspection as a system management unit. Output to.

  The configuration and operation of the entire substrate inspection apparatus 10 have been described above. Subsequently, information used to determine the inspection area will be described with reference to FIGS. 3 to 4B, details of the image processing unit 22 will be described with reference to FIGS. 5 and 6, and FIG. 7 will be referred to. An example of a method for determining detailed conditions for inspection will be described.

  FIG. 3 is an example of the design information 103 included in the manufacturing process / type information mm for inspection. FIG. 4A is a diagram showing a design image 104 that is an image obtained by visualizing the design information 103 and is also an image showing a semiconductor wafer. FIG. 4B is a diagram for explaining the relationship between the design information 103 and the design image 104.

In the example of FIG. 3, the design information 103 includes the following items and indicates the arrangement state of chips on the semiconductor wafer. Details of (7) and (8) among the following items will be described later with reference to FIG. 4B.
(1) “WaferID” is identification information indicating the type of semiconductor wafer.
(2) “ChipSizeX” and “ChipSizeY” are the sizes of the chip formed on the semiconductor wafer in the X and Y directions, respectively.
(3) “DieSizeX” and “DieSizeY” are the sizes in the X and Y directions, respectively, of the dies periodically arranged in the semiconductor wafer.
(4) “ScribeLineWidthX” and “ScribeLineWidthY” are the widths in the X and Y directions of scribe lines existing between adjacent chips, respectively. In this example, the following relationship is established.

ScribeLineWidthX = (DieSizeX-ChipSizeX) / 2
ScribeLineWidthY = (DieSizeY−ChipSizeY) / 2
(5) “ChipNumberInShotX” and “ChipNumberInShotY” are the numbers of chips arranged in the X and Y directions in one shot, respectively. A “shot” is an exposure unit in an exposure apparatus such as a stepper. The example of FIG. 3 represents that 4 (= 2 × 2) chips are included in one shot and 4 chips are exposed in one shot.
(6) “ShotNumberInWaferX” and “ShotNumberInWaferY” are the numbers of shots exposed in the X direction and the Y direction in one semiconductor wafer, respectively. The example of FIG. 3 represents that one semiconductor wafer is exposed by being divided into 48 (= 8 × 6) regions.
(7) “InspectionChipIndex” is data in which index flags indicating whether or not each chip is an object to be inspected are arranged in accordance with a predetermined order of chips. A chip to be inspected is indicated by a value “1”, and a non-inspection target chip is indicated by a value “0”.
(8) “MatrixShiftX” and “MatrixShiftY” are respectively the X direction and the center of the matrix indicating the layout pattern in which a plurality of shots formed corresponding to one semiconductor wafer are arranged, and the center of the semiconductor wafer. Indicates the offset amount in the Y direction.

The values for the items (5), (6), and (8) above are determined by the time of exposure at the latest. However, if it is determined in advance that a specific exposure apparatus is used, the value is determined at the time of design. Sometimes it is decided.
(9) “EdgeCutWidth” represents the edge cut width of the outer periphery of the semiconductor wafer.
(10) “WaferDiameter” represents the diameter of the semiconductor wafer.

  The design image 104 in FIG. 4A is an image obtained by visualizing the design information 103 in FIG. In the present embodiment, the design image 104 is a monochrome luminance image. The design image 104 is generated from the design information 103 in the inspection manufacturing process / product information mm by the design image generation unit 21 described later.

  In the design image 104, individual chips formed on the semiconductor wafer are represented by rectangular chip areas 105, and the chip areas 105 are periodically arranged. In addition, scribe lines provided between adjacent chips to divide individual dies corresponding to the respective chips are represented by linear scribe line regions 106.

  Further, the outer periphery of the semiconductor wafer is represented as a linear edge cut area 107 that draws a circle in the design image 104. A region that is not used as a chip in the peripheral portion of the semiconductor wafer is represented as a chip outside region 108. In other words, the chip outside area 108 is an area that does not belong to any of the chip area 105, the scribe line area 106, and the edge cut area 107. Further, in the design image 104, the shape of the notch serving as a reference for the orientation of the semiconductor wafer is also represented as a notch region 109b.

  As described above, the design image 104 is separated into a plurality of regions having different properties in the semiconductor wafer. It is preferable to perform a macro inspection on each separated region according to the difference in properties. That is, each area in the design image 104 is an inspection area as an area to be inspected under a certain condition. The design image 104 includes a chip area 105, a scribe line area 106, an edge cut area 107, and an inspection area. An off-chip area 108 is included.

  Here, the “inspection area” is an area to be inspected, but there are an inspection area in a substrate such as a semiconductor wafer and an inspection area in an image representing the substrate. The images representing the substrate include a design image 104, an original macro image 100, a macro image 100 processed by translation / rotation conversion, and the like.

  In order to perform a macro inspection by image processing, an area corresponding to the inspection area on the substrate is determined on the original macro image 100 or the processed macro image 100 in advance before the macro inspection, and the inspection on the image is performed. It is necessary to set as an area. The inspection area on the design image 104 is uniquely derived from the design information 103 and used to set the inspection area on the original macro image 100 or the processed macro image 100.

That is, the design information 103 includes the definition of the inspection area on the substrate, and the inspection area on the design image 104 is a visualization of the definition included in the design information 103.
For the convenience of illustration, it cannot be sufficiently distinguished in FIG. 4A, but in the design image 104, the chip area 105, the scribe line area 106, the edge cut area 107, and the chip outside area 108 are different from each other. The brightness is assigned. The assignment is reflected in the design image 104. Therefore, in image processing, any pixel in the design image 104 can easily be determined as to which of the four areas belongs from the assigned luminance.

  In order to assign different luminance to each region, for example, the board inspection apparatus 10 may receive an instruction from the operator via a screen as shown in FIG. Alternatively, the default brightness that should be assigned to each area by a hard disk device or the like may be stored in advance, and the design image generation unit 21 may refer to the default brightness and assign it to each area when the design image 104 is generated.

  FIG. 4B is a diagram showing the relationship between the design information 103 and the design image 104. FIG. 4B is an example in which the values of “ShotNumberInWaferX” and “ShotNumberInWaferY” in the design information 103 are 8 and 6, respectively.

  The shot 122 is a rectangular area that is a unit of layout of a plurality of chips or dies and is also a unit for exposing a circuit pattern. The 48 (= 8 × 6) rectangular shots 122 are arranged so as to include at least a region in which a chip is formed on the semiconductor wafer. A matrix 121 in FIG. 4B shows a two-dimensional layout of shots 122. In FIG. 4B, the outer periphery of the design image 104 of the circular semiconductor wafer is shown so as to overlap with the matrix 121 to help understanding.

  In the example of FIG. 4B, the center of the semiconductor wafer and the center of the matrix 121 do not necessarily match. Therefore, the deviation in the X direction and the Y direction between the center of the semiconductor wafer and the center of the matrix 121 is managed in the design information 103 as offset amounts “MatrixShiftX” and “MatrixShiftY” of the design information 103. Note that the values of “MatrixShiftX” and “MatrixShiftY” may be zero as in the examples of FIGS. 3 and 4B.

  FIG. 4B is an example in which the values of “ChipNumberInShotX” and “ChipNumberInShotY” are both 2 in the design information 103. In FIG. 4B, this is represented by a shot 122 including 4 (= 2 × 2) chip regions 105.

  Therefore, in the design information 103, the entire matrix 121 includes 192 (= 4 × 48) chip regions 105. However, as apparent from FIG. 4B, actually, a part of the 192 chip regions 105 is located outside the circle representing the outer periphery of the semiconductor wafer in the design image 104. Further, some of the 192 chip areas 105 extend over the chip outside area 108 inside the design image 104 and the area outside the design image 104.

  The chip area 105 located outside the design image 104 or outside the chip area 108 is not an inspection target because it is an area where no chip actually exists. Therefore, the “InspectionChipIndex” of the design information 103 associates “1” with the chip area 105 where the chip actually exists and “0” with the chip area 105 where the chip actually does not exist, And non-inspection objects. In other words, “1” indicates validity and “0” indicates invalidity.

  In the example of FIG. 4B, “InspectionChipIndex” is data in which 192 index flags having a value of “1” or “0” are arranged. In addition, 192 index flags are arranged in “InspectionChipIndex” in an order of tracing 192 chip regions 105 in the matrix 121 in a predetermined order such as “from left to right and from top to bottom”.

  Since the relative positional relationship between the matrix 121 and the circular design image 104 is defined by the design information 103, the value of each index flag in the “InspectionChipIndex” is also an item other than “InspectionChipIndex” in the design information 103. Can be determined based on. For example, when the inspection object information bb includes only information of items other than “InspectionChipIndex” in the design information 103, the manufacturing process / product information acquisition unit 12 sets the value of each index flag based on the inspection object information bb. It may be calculated and embedded as part of the design information 103 in the inspection manufacturing process / product information mm.

The information used to determine the inspection area has been described above. Next, details of the image processing unit 22 will be described with reference to FIGS. 5 and 6.
FIG. 5 is a functional configuration block diagram of the image processing unit 22 in the present embodiment. The image processing unit 22 includes a design image generation unit 21, a macro image storage unit 31, a design image storage unit 32, an alignment processing unit 55, a switch 37, and a result information generation unit 39. The alignment processing unit 55 includes a resolution conversion unit 33, a resolution conversion unit 34, a collation processing unit 35, a macro image conversion unit 36, and a resolution designation unit 38. The image processing unit 22 is an example of the inspection area setting unit.

  In another embodiment, the image processing unit 22 may further perform various image processing in the macro inspection based on the macro image 100. For example, the image processing unit 22 may perform various image processing such as defect extraction, defect classification, and chip / semiconductor wafer quality determination.

  However, in the present embodiment, the image processing in the macro inspection is performed by the inspection information generation unit 23, and only the portion directly related to the processing for setting the inspection region on the macro image 100 in the recipe creation before starting the macro inspection. Is shown in FIG.

  Each unit in the image processing unit 22 can be realized by a dedicated hardware circuit, or can be realized by software using a general-purpose computer. For example, the macro image storage unit 31 and the design image storage unit 32 can be realized by a semiconductor memory or a hard disk device provided in a computer, and the other units can be realized by a computer CPU executing a program. .

  The image processing unit 22 stores the inspection image information pp received from the image storage unit 20 of FIG. 1 in the macro image storage unit 31. As described above, the inspection image information pp includes the data of the macro image 100 of FIG.

  The inspection manufacturing process / product information mm output from the manufacturing process / product information acquisition unit 12 in FIG. 1 is input to the design image generation unit 21. The design image generation unit 21 generates a design image 104 representing a semiconductor wafer by visualizing the design information 103 of the semiconductor wafer as a two-dimensional image from the design information 103 included in the manufacturing process / type information mm for inspection. Then, the design image generation unit 21 outputs design image information qq including the data of the design image 104 to the design image storage unit 32.

  The design image generation unit 21 generates a design image 104 with the same resolution as the macro image 100. That is, when one pixel in the macro image 100 corresponds to a range of p [mm] × p [mm] of the substrate, the design image generation unit 21 has a range of p [mm] × p [mm] as one pixel. As described above, the design image 104 is generated based on the numerical value of the design information 103.

  In the present embodiment, the design image generation unit 21 generates the design image 104 so that the image sizes of the macro image 100 and the design image 104 are equal. That is, if the width of the macro image 100 is w [pixel] and the height is h [pixel], the design image generation unit 21 generates the design image 104 having a width of w [pixel] and a height of h [pixel]. To do.

  For example, the image capturing unit 19 in FIG. 1 generates two-dimensional image information kk including information on the resolution and size of the macro image 100 and outputs the two-dimensional image information kk to the image storage unit 20. Information regarding the size may be included in the inspection image information pp and output to the image processing unit 22. In this case, although no particular arrow is shown in FIG. 5, the inspection image information pp may be output to both the design image generation unit 21 and the macro image storage unit 31 of the image processing unit 22.

  Accordingly, the design image generation unit 21 can recognize the resolution and size of the macro image 100 and generate the design image 104 with the same resolution and size as the recognized resolution and size. Alternatively, even when the default values of the resolution and size of the macro image 100 are determined, the design image generation unit 21 can recognize the resolution and size of the macro image 100.

  In addition, as described with reference to FIG. 4A, in the present embodiment, the design image 104 is a monochrome luminance image. In addition, the design image generation unit 21 generates a design image 104 by assigning different luminances to the four areas of the chip area 105, the scribe line area 106, the edge cut area 107, and the outside-chip area 108. For example, the design image generation unit 21 assigns brightness to each area according to the setting via the screen shown in FIG.

  The design image information qq includes at least data of the design image 104. In addition, the design image generation unit 21 may include information on luminance assigned to the four areas in the design image information qq. The design image information qq generated by the design image generation unit 21 is output to and stored in the design image storage unit 32.

  As described above, the inspection image information pp including the data of the macro image 100 is stored in the macro image storage unit 31, and the design image information qq including the data of the design image 104 is stored in the design image storage unit 32. Subsequently, the alignment processing unit 55 performs alignment processing between the macro image 100 and the design image 104.

  In the present embodiment, the alignment process is repeatedly performed while increasing accuracy. The switch 37 determines and controls whether to continue or end the alignment process. In addition, the result information generation unit 39 generates image processing result information rr to be output to the inspection information generation unit 23 after the repetition is completed.

  Next, the alignment processing unit 55 will be described. In the following, for convenience of explanation, an example is given in which the size of the macro image 100 and the design image 104 generated above is 3600 × 3600 pixels, but this is not intended to limit the image size used in the macro inspection.

  The macro image storage unit 31 outputs macro image information aaa including data of the macro image 100 of 3600 × 3600 pixels to the resolution conversion unit 33 and the macro image conversion unit 36. The design image storage unit 32 outputs design image information bbb including data of the design image 104 of 3600 × 3600 pixels to the resolution conversion unit 34 and the result information generation unit 39.

  The macro image information aaa and the design image information bbb may further include information such as an image size. The design image information bbb may be information having the same content as the design image information qq.

  When the alignment processing unit 55 starts processing, the resolution specifying unit 38 outputs the initial value of the conversion magnification represented by the resolution information ggg to both the resolution conversion unit 33 and the resolution conversion unit 34. In the present embodiment, the initial value of the conversion magnification is a value “1/16” indicating the magnification with respect to the original image size of 3600 × 3600 pixels.

  The resolution conversion unit 33 converts the macro image 100 represented by the macro image information aaa to a desired resolution specified by the resolution information ggg, and compares the converted macro image information ccc including the converted image data with the collation processing unit 35. Output to.

  In the above example representing the magnification at which the resolution information ggg is 1/16 times, the converted macro image information ccc includes the data of the macro image 100 reduced to 225 × 225 pixels. In the present embodiment, the converted macro image information ccc also includes information related to the magnification of “1/16 times” indicated by the resolution information ggg or the size of “225 × 225 pixels” of the reduced image. That is, the converted macro image information ccc includes information indicating the reduction magnification of the reduced macro image 100 represented by the converted macro image information ccc in some form.

  Similarly, the resolution conversion unit 34 converts the design image 104 represented by the design image information bbb into a desired resolution specified by the resolution information ggg, and converts the design image information ddd including data of the converted design image 104. Is output to the verification processing unit 35. Similar to the conversion macro image information ccc, the conversion design image information ddd also includes information indicating the reduction ratio of the reduced design image 104 represented by the conversion design image information ddd in some form.

  The collation processing unit 35 performs alignment processing between the reduced macro image 100 and the reduced design image 104 based on the input converted macro image information ccc and converted design image information ddd. That is, the matching processing unit 35 calculates a difference value regarding the position and rotation between the macro image 100 and the design image 104 reduced at the same magnification.

  The difference value represents a relative positional shift and relative rotation between the macro image 100 and the design image 104. In the present embodiment, the positional shift of the macro image 100 when the design image 104 is used as a reference. And the amount of rotation are calculated by the verification processing unit 35 as a difference value.

  For example, the difference value includes a displacement amount (that is, a position correction value) that is a movement amount in the X direction and the Y direction of translational movement, and a rotation angle (that is, a rotation correction value) that represents the amount of rotational movement. The difference value is finally a value calculated for aligning the macro image 100 with the original resolution represented by the inspection image information pp with the design image 104 with the original resolution represented by the design image information qq. .

The difference value is expressed according to a reference coordinate system defined in the substrate inspection apparatus 10. For example, the matching processing unit 35
-The above XY coordinate system is the reference coordinate system,
In the design image 104 indicated by the design image information qq, the center of the semiconductor wafer is the origin,
The horizontal direction of the design image 104 is the X axis and the right direction is the positive direction of the X axis,
The vertical direction of the design image 104 is the Y axis and the upward direction is the positive direction of the Y axis,
・ The positive direction of rotation is counterclockwise.
Based on the assumption, the difference value is calculated as a signed numerical value.

  Since the resolution information ggg is output to both the resolution conversion unit 33 and the resolution conversion unit 34, the resolution of the image represented by the conversion macro image information ccc and the image represented by the conversion design image information ddd match. Yes. Therefore, when performing matching processing between two images, the matching processing unit 35 does not need to consider enlargement and reduction, and can simply calculate a difference value related to position and rotation.

  For the calculation of the difference value, for example, image features such as edge information in the image respectively represented by the conversion macro image information ccc and the conversion design image information ddd are used. Various matching processes using image features are known, and the matching processing unit 35 can calculate a difference value by using an arbitrary image matching process. For example, the matching processing unit 35 can use a technique such as matching with the other image using a plurality of characteristic regions in one image as a template.

  Neither the macro image 100 nor the design image 104 represents a pattern representing a minute alignment mark in each chip. Therefore, the collation processing unit 35 cannot perform the alignment process between the macro image 100 and the design image 104 based on the coordinates of a specific point where the alignment mark is located.

  However, the collation processing unit 35, for example, determines the circular outline of the semiconductor wafer 101 in the macro image 100 whose resolution is converted and the shape of the circular edge cut region 107 of the design image 104 whose resolution is converted. By comparing as a whole, alignment processing of the reduced macro image 100 and the reduced design image 104 is performed. As described above, the matching processing unit 35 performs the alignment process with accuracy according to the resolution of the image.

  The collation processing unit 35 outputs the difference value information as a result of the alignment process to the macro image conversion unit 36 as the difference value information eeee. Note that the difference value information eee of the present embodiment also includes information indicating the reduction magnification so that the amount of translation can be converted according to the magnification of the image. Since the information indicating the reduction ratio is included in the converted macro image information ccc and the converted design image information ddd as described above, the collation processing unit 35 can recognize and include the difference value information eeee.

  Based on the macro image information aaa output from the macro image storage unit 31 and the difference value information eee output from the macro image conversion unit 36, the macro image conversion unit 36 converts the original macro image 100 of 3600 × 3600 pixels. A macro image 100 of 3600 × 3600 pixels subjected to translation / rotation conversion is generated. Further, the macro image conversion unit 36 outputs the translation / rotation converted image information fff including the data of the macro image 100 subjected to the translation / rotation conversion and the contents of the difference value information eeee to the switch 37.

  The switch 37 sets the transmission destination of the received translation / rotation converted image information fff in either the resolution designation unit 38 or the result information generation unit 39. That is, the switch 37 outputs the translation / rotation conversion image information fff to the resolution designation unit 38 in order to continue the repetition in the alignment processing unit 55, or the translation / rotation conversion image information fff as result information to end the repetition. Whether to output to the generation unit 39 is determined. The criteria for determination by the switch 37 will be described later with reference to FIG.

  Note that, when the repetition is continued in the present embodiment, the switch 37 also outputs the translation / rotation conversion image information fff to the apparatus control unit 11 in FIG. The apparatus control unit 11 may display the progress of the process on the display unit 53 based on the received translation / rotation conversion image information fff, and is used for controlling the imaging unit 51 during the inspection of another semiconductor wafer that follows. May be.

  When the switch 37 decides to continue the repetition, it outputs the translation / rotation conversion image information fff to the resolution designation unit 38. The resolution designating unit 38 determines the conversion magnification for the next resolution conversion to be performed based on the resolution of the image represented by the translation / rotation conversion image information fff. Then, the resolution specifying unit 38 generates resolution information ggg including the determined magnification information and the contents of the translation / rotation converted image information fff, and outputs the resolution information ggg to the resolution converting unit 33 and the resolution converting unit 34.

  For example, when the current conversion magnification is 1/16, the resolution designating unit 38 may determine that the conversion magnification of the next resolution conversion to be 1/8. As described above, the translation / rotation converted image information fff includes the difference value information eeee, and the difference value information eeee includes information about the conversion magnification. Therefore, the resolution designating unit 38 sets the new information based on the translation / rotation converted image information fff. A conversion magnification can be determined.

  Based on the resolution information ggg received from the resolution specifying unit 38, the resolution conversion unit 33 converts the macro image 100 subjected to translation / rotation conversion represented by the translation / rotation conversion image information fff included in the resolution information ggg into the resolution information ggg. Convert with the new conversion magnification that represents.

  Then, the resolution conversion unit 33 outputs converted macro image information ccc including data of the converted macro image 100 to the collation processing unit 35. For example, when the new conversion magnification represents a magnification of 1/8, the converted macro image information ccc is converted to 450 × 450 pixels after translation / rotation conversion of the original 3600 × 3600 pixel macro image 100. The data of the macro image 100 is included.

  Similarly, the resolution conversion unit 34 converts the design image 104 represented by the design image information bbb previously received from the design image storage unit 32 into a new conversion magnification represented by the resolution information ggg newly received from the resolution designation unit 38. The converted design image information ddd including the converted design image 104 data is output to the collation processing unit 35.

  In this way, the collation processing unit 35 that has received the converted macro image information ccc from the resolution conversion unit 33 and the converted design image information ddd from the resolution conversion unit 34 again performs alignment processing in the same manner as described above. The same applies to the subsequent processing.

  If the switch 37 determines that the repetition should be terminated, the switch 37 outputs the translation / rotation conversion image information fff to the result information generation unit 39. Upon receiving the translation / rotation conversion image information fff, the result information generation unit 39 integrates the design image information bbb received from the design image storage unit 32 and the translation / rotation conversion image information fff to generate image processing result information rr. Then, the image processing result information rr is output to the inspection information generation unit 23 in FIG. In the present embodiment, the image processing result information rr is the original resolution 3600 × 3600 pixel design image 104 data and the original resolution 3600 × 3600 pixel macro image translated / rotated based on the difference value. 100 data. The image processing result information rr may further include difference value data.

As described above, the inside of the image processing unit 22 has been described with reference to FIG. 5, but the overall operation of the image processing unit 22 is summarized as follows.
That is, the image processing unit 22 receives the inspection image information pp and the inspection manufacturing process / type information mm as input, generates the design image information qq based on the inspection manufacturing process / type information mm, and the inspection image information. Collation processing based on pp and design image information qq is performed, and the result is output to the inspection information generation unit 23 as image processing result information rr.

Next, details of the operation of the inspection information generation unit 23 in FIG. 1 after receiving the image processing result information rr from the image processing unit 22 will be described.
The inspection information generation unit 23, in addition to the image processing result information rr including the data of the macro image 100 translated / rotated based on the difference value, the inspection manufacturing process / product information mm including the design information 103, the macro Inspection image information pp including data of the image 100 is also received. The inspection information generation unit 23 according to the present embodiment generates inspection area information representing an inspection area on the translated / rotated macro image 100.

  That is, the inspection information generation unit 23 performs a process of determining an area on the macro image 100 aligned with the design image 104 by translation / rotation conversion as an inspection area. As shown in FIG. 4A, the design image 104 has a plurality of areas such as a chip area 105 according to a plurality of attributes. The inspection information generation unit 23 sets the position and range of each area in the design image 104 as an inspection area on the macro image 100 aligned by translation / rotation conversion.

  In addition, in order to distinguish a plurality of areas corresponding to a plurality of attributes and perform inspection under different conditions, the substrate inspection apparatus 10 may accept setting of inspection sensitivity for each area via a screen shown in FIG. Good. The inspection sensitivity is set in association with the index for each area. An example of the value representing the inspection sensitivity is a threshold value of the luminance of the image that is necessary when performing defect extraction.

  The inspection information generation unit 23 generates inspection result information ss which is inspection region information indicating which region on the translated / rotated macro image 100 is to be inspected with what inspection sensitivity, and a system management unit (not shown) Output to. The inspection result information ss includes information related to inspection sensitivity for each region.

  Further, in the inspection result information ss, for example, the position and range of each region are indicated by numerical information describing the coordinate value of the region necessary for the inspection, or by image information in the same format as the design image information qq. That is, the inspection result information ss is information that associates the attributes of each region as the inspection region, the position and range of the inspection region, and the inspection sensitivity with each other.

  The system management unit creates a macro inspection recipe based on the inspection result information ss received from the inspection information generation unit 23, and uses the inspection target information bb including the recipe information as the manufacturing process / product information acquisition unit of the substrate inspection apparatus 10. 12 is output.

  Upon receiving the inspection object information bb including recipe information, the manufacturing process / product information acquisition unit 12 outputs, for example, the inspection manufacturing process / product information mm to the inspection information generation unit 23, and performs the macro inspection to the inspection information generation unit 23. May be instructed to execute.

  The inspection information generation unit 23 instructed to execute the macro inspection performs the macro inspection using the translated / rotated macro image 100. The inspection information generation unit 23 acquires the number of detected defects, feature information regarding the size or classification result of defects, a pass / fail judgment result in units of chips or wafers, and the like as inspection results of the macro inspection. The inspection information generation unit 23 may generate and output inspection result information ss representing the inspection result.

  Next, the processing of the image processing unit 22 described above will be described in more detail with reference to FIG. FIG. 6 is a flowchart showing the operation of the image processing unit 22 in this embodiment. The flowchart of FIG. 6 shows a process for determining an inspection area for creating a recipe. The two-dimensional image information kk representing the macro image 100 is already stored in the image storage unit 20 in FIG. 1 at the start of the processing in FIG.

  In step S151, the design image generation unit 21 in FIG. 5 acquires the manufacturing process / product information mm for inspection including the design information 103 of the semiconductor wafer to be inspected from the manufacturing process / product information acquisition unit 12 in FIG. .

  Next, in step S152, the design image generation unit 21 creates the chip area 105, the scribe line area 106, the edge cut area 107, and the outside chip area 108 that constitute the design image 104 based on the contents of the design information 103. Determine the location of each region.

  As shown in FIG. 3, the design information 103 includes information such as chip size, scribe line width, and chip and shot layout. Further, as described above, the resolution and size of the design image 104 are determined to be the same as those of the macro image 100 of FIG. Therefore, the design image generation unit 21 performs a calculation based on the resolution and size of the macro image 100 and the design information 103, and determines in which position of the design image 104 each region should be arranged using the pixel as a position reference.

  Subsequently, in step S153, the design image generation unit 21 assigns different luminance information to each region in order to identify each region in the design image 104, and generates a design image 104 having the same resolution as the macro image 100. .

  In step S154, the design image generation unit 21 outputs the design image information qq to the design image storage unit 32, and the design image storage unit 32 outputs the design image information bbb to the resolution conversion unit 34 and the result information generation unit 39. . In step S154, the macro image storage unit 31 further acquires the inspection image information pp from the image storage unit 20 and outputs the macro image information aaa to the resolution conversion unit 33 and the macro image conversion unit 36. Further, the resolution specifying unit 38 outputs resolution information ggg indicating the magnification of the initial value to the resolution converting unit 33 and the resolution converting unit 34 in step S154.

Hereinafter, steps S155 to S159 and S161 form a repetitive loop, and step S160 is executed when the repetition is completed.
In step S155, the resolution conversion unit 33 performs the resolution conversion process of the macro image 100 based on the same resolution information ggg output from the resolution specifying unit 38 for the alignment process, and the resolution conversion unit 34 performs the design image 104. The resolution conversion process is performed. Also, the resolution conversion unit 33 and the resolution conversion unit 34 output converted macro image information ccc and converted design image information ddd obtained as a result of the resolution conversion process to the collation processing unit 35, respectively.

  In the next step S156, the collation processing unit 35 extracts and acquires image feature information for use in alignment processing for the macro image 100 whose resolution represented by the converted macro image information ccc has been converted. . An example of the image feature information is edge information.

  In step S157, the matching processing unit 35 uses the acquired image feature information to perform alignment processing between the macro image 100 whose resolution has been converted and the design image 104 whose resolution has been similarly converted. The collation processing unit 35 outputs the difference value information ee obtained as a result to the macro image conversion unit 36. As described above, the difference value information eee includes difference values regarding both translational movement and rotational movement.

  Then, in step S158, the macro image conversion unit 36 performs translation / rotation conversion on the macro image information aaa stored in step S154 based on the calculated difference value. Further, the macro image conversion unit 36 outputs the translation / rotation conversion image information fff obtained as a result to the switch 37.

  In the next step S159, the switch 37 determines whether or not to continue the repetition based on the translation / rotation conversion image information fff, particularly based on the accuracy of the difference value indicated by the translation / rotation conversion image information fff. If the difference value calculated in step S157 in the current iteration loop is within a predetermined range, or if the resolution specified in the conversion in step S155 in the current iteration loop is a predetermined value, the switch 37 is It is determined that the repetition should be terminated, and the process proceeds to step S160. Otherwise, the switch 37 determines that the repetition should be continued, and the process proceeds to step S161.

  In step S159, “the difference value is within a predetermined range” means that, for example, the amount of translation in the X direction and the Y direction is within ± 1/2 of the pixel size of the original macro image 100, and It means that the amount is within ± 0.1 °. That is, in this example, the amount of translation in the X direction and the Y direction is within a predetermined range as long as it is within ± 1/2 pixel in terms of the resolution of the original macro image 100 represented by the inspection image information pp. Is recognized.

  Here, numerical values such as “within ± 1/2” and “within ± 0.1 °” are merely examples, and values according to the accuracy required for the macro inspection can be adopted according to the embodiment. In the present embodiment, the difference value is represented by a signed numerical value, so that the “predetermined range” is defined by an upper limit and a lower limit.

  In step S159, “the resolution is a predetermined value” means, for example, that the resolution enlargement / reduction magnification is “1”. This meaning will be described later together with step S161.

  If it is determined in step S159 that the switch 37 should end the repetition, the process proceeds to step S160. In step S160, the switch 37 outputs the translation / rotation conversion image information fff to the result information generation unit 39.

  As described above, the translation / rotation converted image information fff includes the data of the macro image 100 subjected to translation / rotation conversion and the difference value information eeee. The result information generation unit 39 stores the design image information bbb output from the design image storage unit 32 in step S154.

  Therefore, in step S160, the result information generation unit 39 integrates information representing the design image 104 and information regarding the difference value in a predetermined format based on the translation / rotation conversion image information fff and the design image information bbb, and performs image processing. The result information rr is output to the inspection information generation unit 23 in FIG. 1, and the processing in FIG. 6 is terminated.

For example, the information regarding the difference value included in the image processing result information rr is the difference value itself and / or the macro image 100 that is translated / rotated based on the difference value.
If it is determined in step S159 that the switch 37 should continue to repeat, the process proceeds to step S161. In step S161, the switch 37 outputs the translation / rotation conversion image information fff to the apparatus control unit 11 in FIG. 1 and the resolution designation unit 38 in FIG.

  In step S161, the resolution specifying unit 38 further resets the magnification for converting the resolution based on the information about the resolution (in other words, information about the conversion magnification) included in the translation / rotation conversion image information fff, Is output to the resolution conversion unit 33 and the resolution conversion unit 34 as accurate resolution information ggg. After the end of step S161, control returns to step S155.

  For example, the initial value of the magnification represented by the resolution information ggg may be 1/16, and may be set to double every time step S161 is executed. That is, the magnification may be set in order of 1/16, 1/8, 1/4, 1/2, and 1 for each repetition.

  In other words, for example, in an example in which the size of the original macro image 100 is 3600 × 3600 pixels, the size is 225 × 225 pixels, 450 × 450 pixels, 900 × 900 pixels, 1800 × 1800 pixels, 3600 × 3600 pixels. The alignment processing may be performed sequentially using the created macro image 100 and design image 104.

  Depending on the difference value, the process may proceed from step S159 to step S160 when the magnification is smaller than 1. In the present embodiment, as described above, in step S159, “resolution is a predetermined value” means that the magnification / reduction magnification of the resolution is “1”, so a magnification larger than 1 is set. Never happen.

  In this way, starting from a magnification that is greatly reduced to “1/16 times” or the like, and repeating the process while gradually bringing the magnification closer to 1 is, in other words, repeating the process while gradually increasing the accuracy of the alignment process. It is. That is, the present embodiment has a feature that the processing is performed by changing the resolution of the macro image 100 and the design image 104 in accordance with the accuracy of the calculated difference value.

  Step S159 represents that the process is repeated until the accuracy of the difference value reaches a desired accuracy. By executing the alignment process by such repetition, it is possible to calculate the difference value with high accuracy while suppressing the calculation amount. The reason is as follows.

  The translation / rotation difference value calculated by the collation processing unit 35 depends on the resolutions of the macro image 100 and the design image 104 whose resolution has been converted by the resolution conversion unit 33 and the resolution conversion unit 34. The lower the resolution, the larger the area on the subject corresponding to one pixel and the smaller the number of pixels, so the image size is smaller and the accuracy of the difference value is lower. If the accuracy of the difference value is low, the deviation in the positional relationship between the macro image 100 and the design image 104 is large. However, the lower the resolution, the faster the alignment process can be performed.

  Conversely, the higher the resolution, the smaller the area on the subject corresponding to one pixel and the larger the number of pixels, so the image size is larger and the accuracy of the difference value is higher. If the accuracy of the difference value is high, the positional deviation between the macro image 100 and the design image 104 is small. Also, the higher the resolution, the longer the alignment process.

  In the present embodiment, the macro image 100 having a low resolution is used to such an extent that an alignment mark or the like that is positioned and arranged with high accuracy is not captured. In order to perform the alignment process between the macro image 100 and the design image 104 without using an alignment mark or the like, the pattern of the entire image is not compared with the positions of specific points whose positions are strictly determined. Matching is effective.

  The resolution of the original macro image 100 is not so high as to show the alignment mark. However, if an attempt is made to perform alignment using the entire macro image 100 and the design image 104, the resolution of the original macro image 100 is high enough that the influence of the processing time cannot be ignored.

  Therefore, if two low-resolution images are input, even if the entire image is used for matching, it is possible to perform alignment processing at a relatively high speed. The resolution is coarser than the required accuracy. This is because the following effects can be obtained.

  In general, in the alignment process using image feature information, a process for searching where a feature point extracted from one image corresponds to the other image is performed. Therefore, by limiting the image information used for the search, that is, the range of the image to be searched, the calculation amount of the alignment process can be reduced.

  In the present embodiment, coarse alignment processing is first performed with low accuracy using data with low accuracy. As a result, the deviation between the macro image 100 and the design image 104 after roughly coarse alignment is within a certain range according to low accuracy. Therefore, the search range in the subsequent detailed alignment process with higher accuracy is limited, and the processing time of the subsequent alignment process can be reduced.

  Therefore, in the present embodiment, a method is employed in which the alignment process is repeatedly performed a plurality of times while starting with a low resolution and gradually changing to a higher resolution. By combining the alignment processing with a plurality of resolutions, a low resolution merit that the calculation amount is small and the processing time is short, and a very high accuracy difference value are obtained, and the accuracy of the image processing result information rr obtained as a final result is obtained. You can enjoy the advantages of high resolution that is very high.

  In consideration of the trade-off between the calculation amount that is increased by repeating the alignment process a plurality of times and the calculation amount that is decreased by gradually limiting the search range, the initial value of the conversion magnification, and in step S161 It is desirable to determine the degree of change when resetting the conversion magnification.

  In this embodiment, when the conversion magnification becomes 1, the process proceeds from step S159 to step S160. However, an embodiment in which the matching process is repeated even if the conversion magnification is larger than 1 is possible. In such an embodiment, for example, a registration process may be performed between the macro image 100 and the design image 104 enlarged by an appropriate interpolation technique. Further, the magnification information such as 2 times, 4 times, or 8 times may be sequentially specified by the resolution information ggg.

Next, an example of a user interface in the present embodiment will be described with reference to FIG.
FIG. 7 is a screen configuration diagram at the time of operation in the present embodiment. FIG. 7 is an example of the configuration of the screen centering on the location where the macro image 100 and the design image 104 are displayed. In other embodiments, the image display unit 110 may display other information. Further, the macro image 100 in the description of FIG. 7 is the macro image 100 acquired by the imaging unit 51 and the macro image 100 translated / rotated by the image processing unit 22 according to the progress of the processing. There is a case.

  The image display unit 110 in FIG. 7 is a screen displayed on the display unit 53 in FIG. The image display unit 110 is divided into the following four areas (1) to (4), and further includes a button (5).

(1) The area with the heading “Wafer Image” is an area for displaying the macro image 100 together with the superimposed image display box 115 indicating a rectangular range.
For example, upon receiving the image storage control information nn indicating “image information acquisition completion”, the apparatus control unit 11 converts the macro image 100 into the area (1) based on the two-dimensional image information kk stored in the image storage unit 20. Display control may be performed. Further, for example, when the result information generating unit 39 receives the translation / rotation conversion image information fff, the macro image 100 subjected to translation / rotation conversion may be controlled to be displayed in the area (1).

  (2) The area with the heading “Design Image” is an area for displaying the design image 104 together with the superimposed image display box 116 indicating a rectangular range. This area includes a fill display button 111 and a contour display button 112.

  For example, the design image generation unit 21 may generate the design image 104 and perform control for displaying the design image 104 in the area (2). Alternatively, the design image generation unit 21 may output the data of the design image 104 to the device control unit 11, and the device control unit 11 may perform display control of the area (2).

  As described above, different luminance information is assigned to each area such as the chip area 105 in the design image 104. The fill display button 111 is a button for instructing to display each area, which is a closed section, by filling it with the assigned luminance. The contour display button 112 is a button for instructing to display each region by the contour of the assigned luminance.

  When the operator of the board inspection apparatus 10 clicks any button using the input unit 54, the apparatus control unit 11 changes the display method of the design image 104 according to the clicked button, and the design image 104 is displayed by the operator. Are displayed in the desired display method.

  The operator can select an easy-to-see display method from among the fill display and the contour display depending on the substrate to be inspected, so that the state of each region such as the chip region 105 set as the inspection region, particularly the boundary between the regions Can be easily confirmed.

  Further, when the operator changes the size of one of the superimposed image display boxes 115 or 116 by, for example, a drag operation using the input unit 54, the size of the other also changes in conjunction. This is because the device control unit 11 monitors the input from the input unit 54 and controls the display of the superimposed image display boxes 115 and 116.

  Similarly, the relative position of the superimposed image display box 115 in the macro image 100 and the relative position of the superimposed image display box 116 in the design image 104 change and move in conjunction with each other. Thus, under the control of the apparatus control unit 11, the superimposed image display boxes 115 and 116 are displayed in association with each other in position and size.

  (3) The overlapping area 113 with the heading “Overlapped Image” is an area for displaying the macro image 100 and the design image 104 independently or in an overlapping manner. In the overlapping area 113, the macro image 100 and the design image 104 in the ranges specified by the overlapping image display boxes 115 and 116, respectively, are displayed. In addition, the overlapping area 113 includes three selection check boxes for display switching.

  When the operator turns on the check box of the option “Wafer Image” using the input unit 54, the apparatus control unit 11 displays only the macro image 100 in the overlapping region 113.

  When the operator turns on the check box of the option “Design Image” using the input unit 54, the apparatus control unit 11 displays only the design image 104 in the overlapping region 113.

  When the operator turns on the option “Overlapped” check box using the input unit 54, the apparatus control unit 11 displays the macro image 100 and the design image 104 that are processed transparently in the overlapping region 113.

  By superimposing the macro image 100, which is a target for inspection region setting in the recipe creation, and the design image 104 on one overlapping region 113, the contents of the inspection region setting, the macro image 100, The operator can easily check the state of the design information 103.

(4) The inspection information display area 114 with the heading “Information” is an area for displaying information related to the inspection area to be inspected on the macro image 100.
The examination information display area 114 has an input text box “observation magnification” for setting the size of the superimposed image display boxes 115 and 116. “Observation magnification” represents the magnification when the macro image 100 and the design image 104 of the above (1) and (2) are displayed in the overlapping region 113 as an area ratio. When the operator inputs a numerical value in the input text box of “observation magnification” using the input unit 54 and clicks the “set” button on the right side, the apparatus control unit 11 is linked to the input numerical value, The sizes of the superimposed image display boxes 115 and 116 are changed.

The examination information display area 114 includes a plurality of input text boxes for setting the luminance in each area of the design image 104 in association with the attribute of each area of the design image 104.
For example, the attribute of the chip area 105 is “chip”, and the attribute of the scribe line area 106 is “scribe line”. That is, the design information 103 represents a plurality of attributes to be inspected with different inspection sensitivities, and the attributes of each area in the design image 104 are in accordance with the properties of the areas on the substrate represented by the design information 103. It corresponds.

  For example, even if a certain type of defect occurs on the scribe line, it does not affect the product properties of the semiconductor wafer, but if it occurs on the chip, it is necessary to determine the chip as a defective product. This is due to the difference in properties between the chip and the scribe line. This difference in properties is expressed as an attribute.

  In the present embodiment, when the operator inputs the luminance using the input unit 54 and clicks the “set” button, the luminance is a design image via the device control unit 11 that monitors the input from the input unit 54. Input to the generation unit 21. The design image generation unit 21 assigns the input luminance to each area of the design image 104. Then, the design image generation unit 21 performs control for displaying the design image 104 whose appearance has been changed by luminance assignment.

In addition, the design image generation unit 21 inspects the brightness of the plurality of regions so as not to have the same value, and performs processing such as warning the operator when the same brightness value is specified for the plurality of regions.
The examination information display area 114 further includes a plurality of input text boxes for inputting a set of “threshold 1” and “threshold 2” under the heading “detection sensitivity threshold”. Each set corresponds to each region in the design image 104 as in the luminance input text box.

  When the operator inputs numerical values of “threshold 1” and “threshold 2” to each area in the design image 104 using the input unit 54 and clicks the “set” button, the apparatus control unit 11 moves to each area. A set of two threshold values input as parameters representing the corresponding inspection conditions is output to the system management unit, for example, as data used for recipe creation. In the example of FIG. 7, the operator can set two types of threshold values independently for each of the four areas, and can change them as necessary.

  In the macro inspection, image processing for defect extraction and defect classification is performed. For example, the image processing may be performed by the inspection information generation unit 23 of FIG. 1 or may be performed by another unit (not shown) of the substrate inspection apparatus 10. “Threshold 1” and “Threshold 2” have different meanings depending on a specific method of image processing for macro inspection, but are values that define inspection sensitivity for defect extraction.

  For example, in image processing in an embodiment, if the luminance value of a pixel in the chip area 105 is in a range of “threshold 1” or more and “threshold 2” or less associated with the chip area 105, the pixel is normal. Otherwise, it is determined that there is a defect at that pixel location. In another embodiment, “threshold 1” indicates the lower limit of contrast with the periphery for determining a certain region as a defective region, and the lower limit of the size for determining a certain region as a defective region is “ Threshold 2 "may indicate.

  For example, in a certain defect extraction method, by reducing the value of “threshold 1”, defects having a smaller contrast can be detected, and defects can be easily extracted. Increasing the value lowers the detection sensitivity and makes it difficult to extract defects.

  In general, the edge cut area 107 (corresponding to “area 3” in FIG. 7) and the chip outside area 108 (corresponding to “area 4” in FIG. 7) are the chip area 105 (corresponding to “area 3” in FIG. 7). It is sufficient to detect defects with a much lower sensitivity compared to 1). In some cases, it may not be necessary to detect defects in practice.

  Therefore, when a defect extraction method is used in which the detection sensitivity is higher as the value of “threshold 1” is smaller, the operator sets a larger value of “threshold 1” for the edge cut area 107 and the off-chip area 108. Also good. With this setting, the edge cut region 107 and the outside-chip region 108 are practically not subject to inspection, that is, they are masked and not inspected, and defects can not be extracted.

The scribe line is a portion that is not used as an actual circuit product. However, by detecting a defect on the scribe line, it may be possible to detect a defect across two adjacent chips. Therefore, it is desirable to set “threshold 1” and “threshold 2” so that a certain degree of detection sensitivity can be obtained even in the scribe line region 106.
(5) The image display unit 110 includes an examination region determination button 117 on the lower right. The inspection area determination button 117 is used as follows.

  When the imaging unit 51 captures the substrate and acquires the macro image 100, the macro image 100 is displayed in the area (1). Subsequently, when the design image generation unit 21 generates the design image 104, the design image 104 is displayed in the area (2). When the fill display button 111 or the contour display button 112 is clicked or the luminance is set in the inspection information display area 114 in (4), the design image 104 in the area (2) is changed. The above corresponds to the processing up to step S154 in FIG.

Further, the operator inputs a detection sensitivity threshold at an appropriate timing.
Here, when the operator clicks the examination region determination button 117 using the input unit 54, for example, the apparatus control unit 11 monitoring the input from the input unit 54 sends the image processing unit 22 to the step S154 in FIG. To proceed to step S155. Then, the image processing unit 22 executes the processing after step S155 in FIG.

  For example, when the translation information / rotation conversion image information fff is received in step S160, the result information generation unit 39 performs control to switch the display content of the area (1) to the translation / rotation conversion macro image 100.

Subsequently, from the time when the image processing unit 22 outputs the image processing result information rr to the inspection information generation unit 23 until the execution of the macro inspection, the process is automatically performed without requiring any operator input.
That is, an operator who performs an operation using the input unit 54 while looking at the screen of the image display unit 110 does not need to perform actual alignment manually in order to determine the inspection region on the macro image 100. Only the click operation of the region determination button 117 may be performed.

The embodiment of the present invention has been described above with reference to FIGS. Then, the effect of the said embodiment is demonstrated.
In the above embodiment, the region to be inspected is determined using the macro image 100 obtained by imaging the substrate to be inspected and the design information 103 of the substrate to be inspected. Therefore, according to the above embodiment, the inspection apparatus such as the substrate inspection apparatus 10 having only the macro observation system does not need to include the micro observation system for the alignment process.

  For example, if a known technique is used, a method of aligning design information of a semiconductor wafer and a macro image to be inspected using information on a micro image obtained by micro observation is also possible. In this method, the alignment result is developed into information of a macro image obtained by macro observation. That is, coordinate conversion between the micro image and the macro image is performed. As a result, the setting of the inspection area on the macro image is realized.

  In this method, since the macro inspection apparatus having only the macro observation system cannot obtain the image resolution of micro observation, it is necessary to add a micro observation system unnecessary for the macro inspection itself to the macro inspection apparatus.

  Furthermore, in this method, it is necessary to strictly set the alignment (that is, the positional relationship) between the micro observation system and the macro observation system by attaching the micro observation system. Therefore, imaging control in the macro observation system / micro observation system and transport control when transporting the substrate to be inspected between both observation systems require high accuracy and complicated control is required. Therefore, the scale of the macro inspection apparatus is increased, leading to an increase in cost.

  However, the above embodiment has an advantage that the macro inspection apparatus does not need to include a micro observation system. Moreover, according to the said embodiment, an inspection area | region can be set correctly, requiring an operator's manual operation as much as possible.

  Therefore, the above embodiment contributes to automation of recipe creation. In addition, according to the above-described embodiment, it is possible to accurately perform an inspection such as a pass / fail judgment in a short time on a substrate to be inspected without being influenced by an operator's technique. That is, according to the above-described embodiment, it is possible to prevent the inspection from being performed at the location where the inspection should be performed while preventing the inspection from being performed at the location where the inspection is not necessary, and according to the required inspection sensitivity. Accurate inspection can be realized.

As mentioned above, although embodiment of this invention was described in detail, this invention is not restricted to said embodiment, It is possible to implement various deformation | transformation. Some examples are described below.
A first aspect of the deformation relates to an object to be inspected. In the above embodiment, the application to the macro image 100 of the semiconductor wafer 101 has been described. However, the inspection target is not limited to a semiconductor wafer. For example, an embodiment in which the present invention is applied to a macro image obtained by imaging a glass plate on which one or more LCD panels constituting one display surface are arranged is also possible.

  In this case, a design image is generated from the design information representing the layout of the LCD panel on the glass plate in the same manner as in the above-described embodiment, and the macro image obtained by capturing the actual LCD panel and the design image are subjected to alignment processing. It is possible to implement a method such as determining an appropriate inspection area. The design information in this case includes, for example, the number of chamfers as to how many LCD panels are arranged on one glass plate, the arrangement state of the panels as to how many panels are arranged in the X and Y directions, and the LCD panel The effective size, the interval between adjacent panels, the position where the panel driving circuit is disposed, and the like.

  A second aspect of the transformation relates to the content and format of information. For example, the design information 103 may further include other items that do not exist in FIG. 3, and all of the items in FIG. 3 may not be included in the design information 103.

  Various information formats can be arbitrarily determined according to the embodiment. Furthermore, in some embodiments, some of the various types of information exemplified in the above embodiments can be replaced with different contents.

  For example, in the above embodiment, the resolution information ggg in FIG. 5 includes translation / rotation conversion image information fff. However, the above-described embodiment can be modified such that the resolution information ggg includes only information related to the magnification, and the switch 37 outputs the translation / rotation conversion image information fff to the resolution conversion unit 33. In addition, an embodiment in which the content of information and the flow of information are appropriately modified is possible.

  A third aspect of the modification relates to processing distribution. Although FIG. 1 illustrates the substrate inspection apparatus 10 including both the imaging unit 51 and the inspection processing unit 52, the various processes described above can be distributed to a plurality of apparatuses. That is, the apparatus that performs the alignment process only needs to acquire the macro image 100 and the design image 104 by some method.

  For example, the first device images a substrate to be inspected, transmits a macro image 100 as a result of imaging to the second device, and the third device generates a design image 104 from the design information 103. It may be transmitted to the second device. Then, the second device may perform alignment processing based on the macro image 100 and the design image 104 acquired by reception.

  In this case, the second device may execute image processing for macro inspection using the macro image 100 based on the result of the alignment processing. Alternatively, the second device may transmit the macro image 100 and the result of the alignment process to the fourth device, and the fourth device may execute image processing for macro inspection.

  When processing is distributed to a plurality of devices, the order of processing can be changed as appropriate. In the embodiment of FIG. 1, the acquisition of the two-dimensional image information kk by the imaging unit 51 and the acquisition of the inspection target information bb by the inspection processing unit 52 are performed in parallel, but in the embodiment in which the processing is distributed to a plurality of devices, Processing may be performed in a different order or timing from the embodiment of FIG. For example, in the above example, the fourth device can perform the macro inspection at an arbitrary timing independent of the imaging timing of the first device.

Further, the imaging sensor 16 may be an area sensor, and in that case, one imaging may be performed instead of repeating the imaging of the line.
A fourth aspect of the modification relates to a specific method of the alignment process. In the above embodiment, as shown in FIG. 6, the repeated positioning process is performed while gradually increasing the accuracy. However, in another embodiment, the alignment between the original resolution macro image 100 represented by the inspection image information pp and the design image 104 generated at the same resolution as the original resolution macro image 100 is performed as an image. The processing unit 22 may execute directly in one process.

  A fifth aspect of the deformation relates to an image to be set for the inspection area. In the above-described embodiment, the inspection information generation unit 23 sets an inspection region for the macro image 100 subjected to translation / rotation conversion. However, in other embodiments, the inspection information generation unit 23 may perform inverse transformation on each region such as the chip region 105 represented in the design image 104. A range such as the inversely converted chip region 105 is an inspection region on the original macro image 100 represented by the two-dimensional image information kk.

Alternatively, the image processing unit 22 may be modified to translate / rotate the design image 104 instead of the macro image 100.
As described above, the target for which the inspection information generation unit 23 sets the inspection region may be the translated / rotated macro image 100 or the original macro image 100. However, any macro image 100 is the same in that the imaging unit 51 is an image obtained by imaging the substrate.

  That is, an image obtained by two-dimensionally combining the output from the line sensor, an image directly output from the area sensor, an image subjected to correction such as shading correction after imaging, and translation / rotation conversion after imaging. Each of the performed images is a kind of image obtained by imaging the substrate. Which image obtained by imaging the substrate is the target for setting the inspection region may differ depending on the embodiment.

  In addition, the macro image 100 of the above embodiment is an image obtained by imaging a substrate without using a microscope. However, in some embodiments, an image obtained by imaging a substrate using a low-magnification microscope is displayed as a macro. The image 100 may be used. Here, “low magnification” means that only a resolution such that a pattern such as an alignment mark that enables accurate alignment cannot be recognized by image processing is obtained.

  In some embodiments, the macro image 100 may not be an image of the entire substrate. For example, the first half may inspect the right half of the semiconductor wafer, and the second may inspect the left half of the semiconductor wafer. In this case, the macro image 100 and the design image 104 may be images of only the right half or the left half of the semiconductor wafer.

It is a functional block diagram of the board | substrate inspection apparatus in one Embodiment. It is an example of the macro image which the image acquisition part produced | generated. It is an example of the design information contained in the manufacturing process / type information for inspection. It is a figure which shows a design image. It is a figure explaining the relationship between design information and a design image. It is a functional block diagram of the image processing unit in one embodiment. It is a flowchart which shows operation | movement of the image process part in one Embodiment. It is a screen block diagram at the time of operation in one Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Board | substrate inspection apparatus 11 Apparatus control part 12 Manufacturing process / kind information acquisition part 13 Illumination system control part 14 Imaging sensor control part 15 Illumination part 16 Imaging sensor 17 Stage control part 18 Inspection stage 19 Image capture part 20 Image storage part 21 Design Image generation unit 22 Image processing unit 23 Inspection information generation unit 31 Macro image storage unit 32 Design image storage unit 33 Resolution conversion unit 34 Resolution conversion unit 35 Collation processing unit 36 Macro image conversion unit 37 Switch 38 Resolution designation unit 39 Result information generation unit 40 alignment processing unit 51 imaging unit 52 inspection processing unit 53 display unit 54 input unit 55 alignment processing unit 100 macro image 101 semiconductor wafer 102 chip 103 design information 104 design image 105 chip region 106 scribe line region 107 edge cut region 108 chip Outside Area 109a Notch 109b Notch area 110 Image display section 111 Fill display button 112 Outline display button 113 Overlay area 114 Inspection information display area 115 Overlay image display box 116 Overlay image display box 117 Inspection area determination button 121 Matrix 122 Shot aa control Information bb Inspection object information cc Illumination control information dd Imaging sensor control information ee Stage control information ff Illumination information gg Imaging information hh Stage drive information jj Line image information kk Two-dimensional image information mm Manufacturing process / variety information for inspection nn Image storage control information pp image information for inspection qq design image information rr image processing result information ss inspection result information aaa macro image information bbb design image information ccc conversion macro image information ddd conversion design image Distribution eee difference value information fff translational / rotational transformation image information ggg resolution information

Claims (8)

A substrate inspection apparatus for setting a region on a substrate image obtained by imaging a substrate to be inspected as a substrate image inspection region to be inspected,
Image acquisition means for acquiring the substrate image;
Design information acquisition means for acquiring design information of the substrate, including a definition of a substrate inspection region to be inspected in the substrate;
An image generated from the design information that is a design image representing the substrate is acquired, the design image and the substrate image are collated, and the substrate inspection region is handled based on the collation result and the design information. Inspection region setting means for setting the region on the substrate image as the substrate image inspection region;
A board inspection apparatus comprising: The inspection area setting means is
Design image generation means for generating the design image having the same resolution as the substrate image based on the design information;
Alignment processing means for calculating a difference value indicating relative displacement and rotation between the design image and the substrate image,
Based on the region on the design image defined by the design information corresponding to the substrate inspection region and the difference value, the region on the substrate image corresponding to the substrate inspection region is determined, and the determined Setting an area as the substrate image inspection area;
The substrate inspection apparatus according to claim 1.   The substrate inspection apparatus according to claim 2, wherein the alignment processing unit repeatedly calculates the difference value while changing a resolution of the design image and the substrate image. The design information defines a plurality of regions on the substrate corresponding to a plurality of attributes as the substrate inspection regions,
The inspection area setting means, for each attribute,
An area on the design image defined by the design information corresponding to the board inspection area of the attribute is set to a luminance or a color according to the inspection sensitivity associated with the attribute, and is associated with the inspection sensitivity. And setting the board image inspection area corresponding to the board inspection area of the attribute,
The substrate inspection apparatus according to claim 2. Display means for displaying the substrate image and the design image independently or in an overlapping manner,
The board inspection apparatus according to claim 2, further comprising:   The area on the design image defined by the design information corresponding to the board inspection area is painted with brightness or color corresponding to the board inspection area, or expressed with an outline with the brightness or color. The substrate inspection apparatus according to claim 5, wherein the display unit displays the design image.   The substrate inspection apparatus according to claim 1, wherein the substrate image is an image captured at a low resolution such that an alignment mark for alignment provided on the substrate is indistinguishable. A method in which a computer sets a region on a substrate image obtained by imaging a substrate to be inspected as a substrate image inspection region to be inspected,
Acquiring the substrate image;
Including the definition of a substrate inspection area to be inspected in the substrate, obtaining design information of the substrate,
Obtaining an image generated from the design information and representing the substrate;
Collating the design image with the substrate image,
Based on the collation result and the design information, the region on the substrate image corresponding to the substrate inspection region is set as the substrate image inspection region.
An inspection area setting method characterized by the above.