JP2021196326A - Screen mask inspection device - Google Patents

Abstract

To provide a screen mask inspection device capable of suppressing the occurrence of printing defects or the like.SOLUTION: A metal mask inspection device 30 includes: lighting devices 32A and 32B that can radiate predetermined light to a predetermined portion to be inspected, which is a peripheral portion of an opening 21 on a back surface 201 side, which is a board contact surface side that comes into contact with a printed circuit board during solder printing among both surfaces of the front and back of a metal mask 20; a camera 32C that takes an image of the portion to be inspected of the back surface 201 of the metal mask 20; and a control device that controls those. The control device is configured to acquire shape data on the portion to be inspected based on image data on the portion to be inspected captured by the camera 32C, and to be able to determine based on the shape data on the portion to be inspected whether the portion to be inspected is good or not.SELECTED DRAWING: Figure 4

Description

The present invention relates to a screen mask inspection apparatus for inspecting a screen mask used when performing solder printing on a substrate.

Generally, in a production line in which electronic components are mounted on a substrate, cream solder is first printed on a land of the substrate by screen printing in a solder printing machine. Next, the electronic component is temporarily fixed on the substrate based on the viscosity of the cream solder. After that, soldering is performed by guiding the substrate to the reflow furnace.

The solder printing machine is provided with a screen mask in which a large number of openings are formed in advance corresponding to the land arrangement of the substrate to be printed. In solder printing, the screen mask is in contact with the surface of the substrate, and cream solder is supplied onto the screen mask to slide the squeegee to push the cream solder into the opening. After that, by separating the screen mask from the substrate, the cream solder filled in the opening is ejected and printed (transferred) on the land.

In such a solder printing machine, the screen mask is replaced, for example, when a solder printing defect occurs or when the type of the substrate to be printed is switched.

Then, when a new solder printing is started using the replaced screen mask, the screen mask may be inspected in advance in order to prevent the occurrence of printing defects and the like.

As a technique for performing such an inspection, for example, a technique for determining the quality of a screen mask by three-dimensionally measuring the upper surface of the screen mask with a laser measuring device while the screen mask is attached to a solder printing machine is known. (See, for example, Patent Document 1).

Japanese Unexamined Patent Publication No. 2001-315310

However, in the configuration according to Patent Document 1, of the front and back surfaces of the screen mask, the substrate non-contact surface side opposite to the back surface (lower surface) side, which is the substrate contact surface side that abuts on the printed circuit board during solder printing. It is configured to perform three-dimensional measurement from the surface (upper surface) side.

Therefore, it may not be possible to detect various abnormalities (defective portions) on the back surface side of the screen mask, which may cause bleeding or sagging of the cream solder transferred onto the land during solder printing.

For example, when the peripheral corner of the opening on the back surface side of the screen mask is too rounded due to excessive electrolytic polishing treatment or the like [see FIG. 10 (b)], or the dent or scraping connected to the opening on the back surface side of the screen mask. If there are scratches such as [see FIG. 10 (c)], or if the peripheral edge of the opening on the back surface side of the screen mask is distorted [see FIG. 10 (d)], there will be scratches during solder printing. Cream solder may flow from the screen and cause bleeding or sagging. As a result, printing defects such as bridges are likely to occur, and the mounted parts may not be properly soldered, resulting in an increase in the occurrence rate of defective products.

The present invention has been made in view of the above circumstances and the like, and an object of the present invention is to provide a screen mask inspection apparatus capable of suppressing the occurrence of printing defects and the like.

Hereinafter, each means suitable for solving the above-mentioned problems will be described separately for each item. In addition, the action and effect peculiar to the corresponding means will be added as necessary.

Means 1. A screen mask inspection device for inspecting a screen mask that has a plurality of openings and is used when performing solder printing on a substrate.
Of the front and back surfaces of the screen mask, at least a predetermined light can be applied to a predetermined portion to be inspected, which is a peripheral portion of the opening on the back surface side which is the substrate contact surface side which comes into contact with the substrate during solder printing. One irradiation means and
An image pickup means capable of taking an image of the part to be inspected irradiated with the predetermined light, and an image pickup means.
A shape data acquisition means capable of acquiring shape data related to the inspected portion based on one or a plurality of image data of the inspected portion imaged by the image pickup means.
A screen mask inspection apparatus comprising: a determination means capable of determining the quality of the inspected portion based on the shape data related to the inspected portion acquired by the shape data acquisition means.

According to the above means 1, of the front and back surfaces of the screen mask, the inspection is performed on the predetermined inspected portion, which is the peripheral portion of the opening, from the back surface side, which is the substrate contact surface side that comes into contact with the substrate during solder printing. be able to.

This makes it possible to detect various abnormalities (defective parts) related to the vicinity of the opening (inspected portion) on the back surface side of the screen mask, which may cause solder bleeding or sagging during printing. As a result, it is possible to suppress the occurrence of printing defects and the like.

Means 2. The determination means is
It is determined whether or not the size of the chamfered portion on the periphery of the opening exceeds a predetermined amount (for example, whether or not at least one of the "depth" and "width" of the chamfered portion exceeds a predetermined threshold value). The screen mask inspection apparatus according to means 1, wherein the quality of the inspected portion can be determined by the screen mask inspection apparatus.

As described in the above "Problems to be Solved by the Invention", when the peripheral corner portion of the opening on the back surface side of the screen mask is too rounded due to excessive electrolytic polishing treatment or the like [see FIG. 10 (b)]. There is a risk that solder will flow from there during solder printing, causing bleeding and sagging. On the other hand, according to the above means 2, it is possible to prevent the occurrence of such a problem.

Means 3. The determination means is
Whether or not the inspected portion has an abnormal portion exceeding a predetermined size (for example, at least one of "depth", "width", "length" and "volume" determines a predetermined threshold value. The screen mask inspection apparatus according to means 1 or 2, wherein the quality of the inspected portion can be determined by determining (whether or not there is an abnormal portion exceeding).

As described in the above-mentioned "Problems to be Solved by the Invention", when there are scratches such as dents and scrapes connected to the opening on the back surface side of the screen mask [see FIG. 10 (c)], or the screen mask. If the peripheral edge of the opening on the back surface side of the screen is distorted [see FIG. 10 (d)], cream solder may flow from the peripheral edge of the opening during solder printing, which may cause bleeding or sagging. On the other hand, according to the above means 3, it is possible to prevent the occurrence of such a problem. That is, the above-mentioned "abnormal portion" includes scratches such as dents and scrapes connected to the opening, and distortion of the peripheral edge of the opening. Further, the above-mentioned "abnormal portion" may include a portion to which a foreign substance adheres (foreign matter adhered to the inspected portion) or the like.

Means 4. For a neural network having a coding unit (encoder) that extracts a feature amount from input shape data and a decoding unit (decoder) that reconstructs shape data from the feature amount, the inspected unit has no abnormality. It is equipped with an identification means (generation model) generated by training only the relevant shape data (only the shape data related to non-defective products) as training data.
The determination means is
A reconstructed shape capable of acquiring reconstructed shape data related to the inspected portion reconstructed by inputting the shape data related to the inspected portion acquired by the shape data acquisition means into the identification means as original shape data. Data acquisition method and
A comparison means capable of comparing the original shape data and the reconstructed shape data is provided.
The screen mask inspection apparatus according to any one of means 1 to 3, wherein the quality of the inspected portion can be determined based on the comparison result by the comparison means.

The above-mentioned "neural network" includes, for example, a convolutional neural network having a plurality of convolutional layers. Further, the above-mentioned "learning" includes, for example, deep learning. The above-mentioned "identification means (generation model)" includes, for example, an autoencoder (self-encoder), a convolutional autoencoder (convolutional self-encoder), and the like.

According to the above means 4, it is determined whether or not there is an abnormality (defective part) in the inspected portion by using an identification means (generation model) such as a self-encoder constructed by learning a neural network. .. This makes it possible to detect minute abnormalities that were difficult to detect in the past.

Further, in this means, since the original shape data obtained by imaging the inspected portion and the reconstructed shape data obtained by reconstructing the original shape data are compared, both shape data to be compared are compared. Based on the difference in the imaging conditions on the screen mask side (for example, the position and angle of the screen mask, the deflection, etc.) and the imaging conditions on the inspection device side (for example, the lighting state and the angle of view of the camera), which are the objects to be inspected. It has no effect and enables more accurate detection of finer abnormalities.

In addition, when inspecting a predetermined inspected portion (peripheral portion of the opening) in a predetermined position of the screen mask, the configuration information (position data, dimensional data, etc.) of the inspected portion is used as a quality judgment criterion. In a configuration that requires, for example, substrate design information such as Gerber data is stored in advance, the configuration information of the part to be inspected is appropriately acquired, and the configuration information to be inspected is compared with the configuration information. Since the quality of the part is judged, the inspection efficiency may decrease. In addition, it is necessary to accurately position the screen mask to the inspection position.

On the other hand, according to this means, since each inspected portion is inspected by using an identification means such as a self-encoder, individual configuration information of each of a large number of inspected portions can be obtained. Since it is not necessary to store it in advance and it is not necessary to refer to it at the time of inspection, it is possible to improve the inspection efficiency.

Means 5. The irradiation means is configured to be capable of irradiating light for three-dimensional measurement (for example, pattern light having a striped light intensity distribution) as the predetermined light.
The shape data acquisition means 1 to be characterized in that the shape data acquisition means is configured to be able to acquire three-dimensional shape data related to the inspected portion by using a predetermined three-dimensional measurement method (for example, a phase shift method). The screen mask inspection apparatus according to any one of 4.

Here, as an example of the above-mentioned "three-dimensional measurement method", a phase shift method for acquiring three-dimensional shape data based on a plurality of image data captured under a plurality of pattern lights having different phases is given. Can be done.

According to the above means 5, whether or not there is an abnormality (defective part) in the inspected portion by acquiring the three-dimensional shape data related to the inspected portion by using a three-dimensional measurement method such as a phase shift method. Can be accurately determined. As a result, the inspection accuracy can be improved.

In addition, during solder printing, cream solder flows from the gap around the opening on the back side of the screen mask, and in order to properly grasp the abnormal part that may cause bleeding or sagging of the cream solder transferred onto the land. Among the shape data, it is important to acquire information (height data) in the height direction of the abnormal portion that can cause the gap.

In this regard, by acquiring the three-dimensional shape data related to the inspected portion as in the present means, it is accurately determined whether or not the inspected portion has an abnormal portion that can cause a gap as described above. be able to. As a result, the inspection accuracy can be improved.

Means 6. The screen mask is arranged so that the back surface side is the lower side.
The irradiation means is arranged so as to be able to irradiate the predetermined light from below with respect to the back surface side of the screen mask.
The screen mask inspection apparatus according to any one of means 1 to 5, wherein the image pickup means is arranged so that the back surface side of the screen mask can be imaged from below.

If the screen mask is arranged so that the back surface side (board contact surface side) is on the upper side and the inspection is performed from above, the inspection is stable due to the influence of ambient light such as indoor lighting. May be difficult to do.

On the other hand, according to the means 6, it is possible to suppress the occurrence of the above-mentioned trouble and improve the inspection accuracy. In particular, it is more effective when the inspection is performed in an environment susceptible to minute changes in luminance, such as when the phase shift method is used.

It is a partially enlarged plan view which enlarged a part of the printed circuit board. It is a block diagram which shows the structure of the production line of a printed circuit board. It is a schematic diagram for demonstrating the printing operation by a solder printing machine. It is a schematic block diagram which shows typically the metal mask inspection apparatus. It is a block diagram which shows the functional structure of a metal mask inspection apparatus. It is a schematic diagram for demonstrating the structure of a neural network. It is a flowchart which shows the flow of the learning process of a neural network. It is a flowchart which shows the flow of an inspection process. (A) is a partially enlarged plan view of a metal mask showing an opening and its peripheral portion viewed from the surface side, and (b) is a sectional view taken along line AA of (a). (A) is a partially enlarged cross-sectional view of a metal mask showing a normal opening and its peripheral portion, and (b) to (e) are portions of a metal mask showing an abnormal opening and its peripheral portion. It is an enlarged sectional view.

Hereinafter, an embodiment of the present invention will be described. First, the configuration of a board on which solder is printed and components are mounted will be described. FIG. 1 is a partially enlarged plan view of a part of a printed circuit board as a substrate.

As shown in FIG. 1, the printed circuit board 1 has a wiring pattern (not shown) made of copper foil and a plurality of lands 3 formed on the surface of a flat plate-shaped base board 2 made of glass epoxy resin or the like. .. Further, the surface of the base substrate 2 is coated with a resist film 4 on a portion other than the land 3. Then, as shown in FIGS. 1 and 3, a viscous cream solder 5 is printed on the land 3. In FIGS. 1 and 3, for convenience, a scattered spot pattern is attached to the portion showing the cream solder 5.

Next, a manufacturing line (manufacturing process) for manufacturing the printed circuit board 1 will be described with reference to FIG. FIG. 2 is a block diagram showing the configuration of the production line 10 of the printed circuit board 1. The production line 10 in the present embodiment is set so that the printed circuit board 1 is conveyed from left to right when viewed from the front side thereof.

As shown in FIG. 2, a solder printing machine 12, a solder printing inspection device 13, a component mounting machine 14, and a reflow device 15 are installed in order from the upstream side (left side of FIG. 2) of the production line 10.

The solder printing machine 12 is for performing a solder printing step of printing cream solder 5 on the land 3 of the printed circuit board 1. As shown in FIG. 3, the solder printing machine 12 has a flat metal mask 20 in which a plurality of openings 21 are formed corresponding to each land 3 on the printed circuit board 1, and a surface 202 of the metal mask 20. It is provided with a squeegee 22 that slides along the line. The metal mask 20 corresponds to the screen mask in this embodiment.

The metal mask 20 (see FIG. 4) in the present embodiment holds a rectangular thin plate-shaped metal mask main body 20a in which the plurality of openings 21 are formed, and four sides of the peripheral edge of the mask main body 20a. It is composed of a rectangular frame-shaped metal frame portion 20b. By holding the thin plate-shaped mask body 20a on the frame portion 20b having high rigidity in a stretched state, it is possible to suppress bending of the mask body 20a.

Under the above configuration, in the solder printing process, first, the metal mask 20 is aligned and brought into contact with the surface side of the printed circuit board 1. Next, the cream solder 5 is supplied to the surface 202 of the metal mask 20. Subsequently, the squeegee 22 is slid along the surface 202 of the metal mask 20 to push the cream solder 5 into the opening 21.

After that, by separating the metal mask 20 from the printed circuit board 1, the cream solder 5 filled in the opening 21 is ejected from the back surface 201 side and printed (transferred) on the land 3 to complete the solder printing. ..

The solder print inspection device 13 is for performing a solder print inspection step for inspecting the state (for example, printing position, height, amount, etc.) of the cream solder 5 printed as described above.

The component mounting machine 14 is for performing a component mounting process of mounting an electronic component 25 (see FIG. 1) on a land 3 on which a cream solder 5 is printed. The electronic component 25 includes a plurality of electrodes and leads, and each of the electrodes and leads is temporarily fixed to a predetermined cream solder 5.

The reflow device 15 is for performing a reflow process in which the cream solder 5 is heated and melted to solder-bond (solder) the land 3 and the electrodes and leads of the electronic component 25.

In addition, although not shown, the production line 10 is provided with a conveyor or the like for transferring the printed circuit board 1 between the above-mentioned devices such as between the solder printing machine 12 and the solder printing inspection device 13. There is. Further, a branching device is provided between the solder printing inspection device 13 and the component mounting machine 14. Then, the printed circuit board 1 determined to be non-defective by the solder printing inspection device 13 is guided to the component mounting machine 14 on the downstream side thereof, while the printed circuit board 1 determined to be defective is directed to the defective product storage unit by the branching device. It will be discharged.

Further, a cleaning device 29 and a metal mask inspection device 30 are installed in the vicinity of the solder printing machine 12. The metal mask inspection device 30 constitutes the screen mask inspection device according to the present embodiment.

The cleaning device 29 is for performing a cleaning step of cleaning the metal mask 20 to which dirt or the like is attached in the solder printing machine 12. In the present embodiment, in the manufacturing process (manufacturing line 10) of the printed circuit board 1 described above, every time the solder printing is performed by the solder printing machine 12 a predetermined number of times, the solder printing is temporarily stopped and used for the solder printing. It is configured to clean the metal mask 20 that has been printed.

The metal mask inspection device 30 inspects the metal mask 20 before being attached to the solder printing machine 12, such as the metal mask 20 that has been cleaned by the cleaning device 29 as described above and the newly replaced unused metal mask 20. It is for doing.

Here, the configuration of the metal mask inspection apparatus 30 will be described in detail with reference to FIGS. 4 and 5. FIG. 4 is a schematic configuration diagram schematically showing the metal mask inspection device 30. FIG. 5 is a block diagram showing a functional configuration of the metal mask inspection device 30.

The metal mask inspection device 30 includes a transport mechanism 31 for transporting and positioning the metal mask 20, an inspection unit 32 for inspecting the metal mask 20, drive control of the transport mechanism 31 and the inspection unit 32, and metal. It includes a control device 33 that executes various controls, image processing, and arithmetic processing in the mask inspection device 30.

The conveyor belt 31 includes a pair of conveyor belts 31a arranged along the loading / unloading direction of the metal mask 20, an endless conveyor belt 31b rotatably arranged for each conveyor belt 31a, and the conveyor belt 31b. It is provided with a driving means (not shown) such as a driving motor and a chuck mechanism (not shown) for positioning the metal mask 20 at a predetermined position, and is driven and controlled by a control device 33 (conveyor mechanism control unit 79 described later). ..

Under the above configuration, in the metal mask 20 carried into the metal mask inspection device 30, the frame portions 20b at both side edges in the width direction orthogonal to the carry-in / out direction are inserted into the transport rails 31a, respectively, and on the conveyor belt 31b. It is placed in. Subsequently, the conveyor belt 31b starts operation, and the metal mask 20 is conveyed to a predetermined inspection position. When the metal mask 20 reaches the inspection position, the conveyor belt 31b stops and the chuck mechanism operates. By the operation of this chuck mechanism, the conveyor belt 31b is pushed up, and the frame portions 20b at both side edges of the metal mask 20 are sandwiched between the conveyor belt 31b and the upper side portions of the transport rail 31a. As a result, the metal mask 20 is positioned and fixed at the inspection position. When the inspection is completed, the fixing by the chuck mechanism is released and the conveyor belt 31b starts operation. As a result, the metal mask 20 is carried out from the metal mask inspection device 30. Of course, the configuration of the transport mechanism 31 is not limited to the above-mentioned embodiment, and other configurations may be adopted.

The inspection unit 32 is arranged below the transport rail 31a (the transport path of the metal mask 20). The inspection unit 32 is the first as an irradiation means for irradiating a predetermined inspection range of the back surface 201 of the metal mask 20 with predetermined light for three-dimensional measurement (patterned light having a striped light intensity distribution) from diagonally below. The lighting device 32A and the second lighting device 32B, the camera 32C as an imaging means for capturing a predetermined inspection range of the back surface 201 of the metal mask 20 from directly below, and the camera 32C capable of moving in the X-axis direction (left-right direction in FIG. 4). The X-axis moving mechanism 32D (see FIG. 5) and the Y-axis moving mechanism 32E (see FIG. 5) that enable movement in the Y-axis direction (see FIG. 4) are provided and are driven and controlled by the control device 33. To.

In this embodiment, the back surface 201, which is the substrate contact surface side of the metal mask 20 that abuts on the printed circuit board 1 during solder printing, faces downward (becomes the lower surface) and is on the substrate non-contact surface side. The metal mask 20 is set with respect to the metal mask inspection device 30 so that the surface 202 facing upward (becomes the upper surface).

Further, the "inspection range" of the back surface 201 of the metal mask 20 is one of a plurality of areas preset on the back surface 201 of the metal mask 20 with the size of the imaging field of view (imaging range) of the camera 32C as one unit. It is an area.

The control device 33 (movement mechanism control unit 76) drives and controls the X-axis movement mechanism 32D and the Y-axis movement mechanism 32E to position and fix the inspection unit 32 at the inspection position. It is possible to move to a position below the inspection range of. Then, the inspection unit 32 is sequentially moved to a plurality of inspection ranges set on the back surface 201 of the metal mask 20, and the inspection related to the inspection range is executed to inspect the entire back surface 201 of the metal mask 20. It is configured to be executed.

The first lighting device 32A forms a first light source 32Aa that emits predetermined light and a first lattice that converts light from the first light source 32Aa into a first pattern light having a striped light intensity distribution. It is provided with a liquid crystal shutter 32Ab and is driven and controlled by a control device 33 (light control unit 72 described later).

The second lighting device 32B forms a second light source 32Ba that emits predetermined light and a second lattice that converts the light from the second light source 32Ba into a second pattern light having a striped light intensity distribution. It is provided with a liquid crystal shutter 32Bb and is driven and controlled by a control device 33 (light control unit 72 described later).

Under the above configuration, the light emitted from each of the light sources 32Aa and 32Ba is guided to a condenser lens (not shown), and after being made into parallel light there, the projection lens (not shown) is passed through the liquid crystal shutters 32Ab and 32Bb. It will be projected as pattern light on the metal mask 20. Further, in the present embodiment, the switching control of the liquid crystal shutters 32Ab and 32Bb is performed so that the phase of each pattern light is shifted by a quarter pitch.

By using the liquid crystal shutters 32Ab and 32Bb as the lattice, it is possible to irradiate a pattern light close to an ideal sine wave. This improves the measurement resolution of the three-dimensional measurement. Further, the phase shift control of the pattern light can be electrically performed, and the device can be made compact.

The camera 32C includes an image pickup element such as a CCD (Charge Coupled Device) type image sensor or a CMOS (Complementary Metal Oxide Semiconductor) type image sensor, and an optical system (lens unit or lens unit) that forms an image of a metal mask 20 on the image pickup element. It has a diaphragm, etc.), and its optical axis is arranged along the vertical direction (Z-axis direction). Of course, the image pickup element is not limited to these, and other image pickup elements may be adopted.

The camera 32C is driven and controlled by the control device 33 (camera control unit 73 described later). More specifically, the control device 33 executes the image pickup process by the camera 32C while synchronizing with the irradiation process by both the lighting devices 32A and 32B. As a result, of the light emitted from the first lighting device 32A or the second lighting device 32B, the light reflected by the metal mask 20 is captured by the camera 32C, and image data is generated.

The image data captured and generated by the camera 32C in this way is converted into a digital signal inside the camera 32C, and then transferred to the control device 33 (image acquisition unit 74 described later) in the form of a digital signal for storage. Will be done. Then, the control device 33 (data processing unit 75 and the like described later) performs various image processing and arithmetic processing described later based on the image data.

The control device 33 temporarily stores a CPU (Central Processing Unit) that executes predetermined arithmetic processing, a ROM (Read Only Memory) that stores various programs, fixed value data, and the like, and various data when executing various arithmetic processing. It consists of a computer including a RAM (Random Access Memory) and peripheral circuits thereof.

When the CPU operates according to various programs, the control device 33 includes a main control unit 71, a lighting control unit 72, a camera control unit 73, an image acquisition unit 74, a data processing unit 75, and a movement mechanism control unit 76, which will be described later. It functions as various functional units such as a learning unit 77, an inspection unit 78, and a transport mechanism control unit 79.

However, the various functional units are realized by the cooperation of various hardware such as the CPU, ROM, and RAM, and it is not necessary to clearly distinguish the functions realized by hardware or software. , Some or all of these functions may be realized by a hardware circuit such as an IC.

Further, the control device 33 includes an input unit 55 composed of a keyboard, mouse, touch panel, etc., a display unit 56 having a display screen such as a liquid crystal display, and a storage unit 57 capable of storing various data, programs, calculation results, and the like. , A communication unit 58 capable of transmitting and receiving various data to and from the outside is provided.

Here, the various functional units constituting the control device 33 will be described in detail. The main control unit 71 is a functional unit that controls the entire metal mask inspection device 30, and is configured to be able to transmit and receive various signals to other functional units such as the lighting control unit 72 and the camera control unit 73.

The lighting control unit 72 is a functional unit that drives and controls the first lighting device 32A and the second lighting device 32B.

The camera control unit 73 is a functional unit that drives and controls the camera 32C, and controls the imaging timing and the like based on the command signal from the main control unit 71.

The image acquisition unit 74 is a functional unit for capturing image data captured and acquired by the camera 32C.

The data processing unit 75 is a functional unit that performs predetermined image processing on the image data captured by the image acquisition unit 74 and performs three-dimensional measurement processing using the image data. For example, in the learning process described later, the learning shape data (three-dimensional shape data for learning) which is the learning data used for learning the deep neural network 90 (hereinafter, simply referred to as “neural network 90”; see FIG. 6)). To generate. Further, in the inspection process described later, inspection shape data (three-dimensional shape data for inspection) is generated.

The movement mechanism control unit 76 is a functional unit that drives and controls the X-axis movement mechanism 32D and the Y-axis movement mechanism 32E, and controls the position of the inspection unit 32 based on the command signal from the main control unit 71.

The learning unit 77 is a functional unit that learns the neural network 90 using learning data or the like and constructs an AI (Artificial Intelligence) model 100 as an identification means.

As will be described later, the AI model 100 in the present embodiment is a neural network using only the shape data related to the inspected portion EA [see FIG. 10 (a) and the like] of the unused metal mask 20 having no abnormality as learning data. It is a generative model constructed by deep learning the 90, and has a so-called autoencoder (self-encoder) structure.

Here, the structure of the neural network 90 will be described with reference to FIG. FIG. 6 is a schematic diagram conceptually showing the structure of the neural network 90. As shown in FIG. 6, the neural network 90 reconstructs the encoder unit 91 as a coding unit for extracting the feature amount (latent variable) TA from the input shape data GA, and the shape data GB from the feature amount TA. It has a structure of a convolutional auto-encoder (CAE) having a decoder unit 92 as a decoding unit.

Since the structure of the convolution autoencoder is known, detailed description thereof will be omitted, but the encoder unit 91 has a plurality of convolution layers 93, and each convolution layer 93 has a plurality of filters for input data. The result of the convolution operation using (Kernel) 94 is output as the input data of the next layer. Similarly, the decoder unit 92 has a plurality of deconvolution layers 95, and each deconvolution layer 95 is the result of performing a deconvolution operation using a plurality of filters (kernels) 96 on the input data. Is output as input data for the next layer. Then, in the learning process described later, the weights (parameters) of the filters 94 and 96 are updated.

The inspection unit 78 is a functional unit that inspects whether or not the inspected portion EA of the metal mask 20 has an abnormality. In the present embodiment, as shown in FIGS. 9A and 9B, the peripheral portion of one opening 21 on the back surface 201 side, which is the substrate contact surface side of the metal mask 20, is one inspected portion. Set as EA. That is, a plurality of inspected portions EA are set on the back surface 201 of the metal mask 20. 9 (a) is a partially enlarged plan view of the metal mask 20 showing the opening 21 and its peripheral portion as viewed from the surface 202 side, and FIG. 9 (b) is a partially enlarged plan view of FIG. 9 (a). It is a sectional view. The peripheral portion (surface 202) of the opening 21 in FIG. 9A is provided with a scattered dot pattern in order to make it easy to discriminate the range.

The transport mechanism control unit 79 is a functional unit that drives and controls the transport mechanism 31, and controls the position of the metal mask 20 based on a command signal from the main control unit 71.

The storage unit 57 is composed of an HDD (Hard Disk Drive), an SSD (Solid State Drive), or the like, and has, for example, a predetermined storage area for storing the AI model 100 (neural network 90 and learning information acquired by learning the neural network 90). ing. Such a storage area constitutes the model storage means in the present embodiment.

The communication unit 58 is provided with a wireless communication interface conforming to a communication standard such as a wired LAN (Local Area Network) or a wireless LAN, and is configured to be able to transmit and receive various data to and from the outside.

Next, the learning process of the neural network 90 performed by the metal mask inspection device 30 will be described with reference to the flowchart of FIG. 7.

First, the operator prepares an unused metal mask 20 having no abnormality (not having various abnormality portions Q described later). The metal mask 20 prepared here preferably has an opening 21 having the same shape as the metal mask 20 to be inspected. However, the thickness and material of the metal mask 20, the size of the opening 21, the arrangement layout, and the like do not need to be the same, and it is preferable to learn based on various types of learning data in terms of versatility.

Then, the operator arranges the unused metal mask 20 having no abnormality at the predetermined inspection position of the metal mask inspection device 30, and then causes the main control unit 71 to execute the predetermined learning program.

When the learning process is started based on the execution of the predetermined learning program, the main control unit 71 first performs preprocessing for learning the neural network 90 in step S101.

In this pre-processing, the same processing as the image data acquisition processing (step S301) and the three-dimensional shape data acquisition processing (step S302) related to the inspection processing described later is executed. Since the details of these processes will be described later, they will be briefly described here.

First, for a predetermined inspection range on the back surface 201 side of the metal mask 20, the phase of the first pattern light emitted from the first lighting device 32A is changed, and four times of imaging are performed under the first pattern light having different phases. After the processing, while changing the phase of the second pattern light emitted from the second lighting device 32B, the imaging process is performed four times under the second pattern light having a different phase, and a total of eight types of area images are imaged. Get the data.

After that, a three-dimensional shape of a predetermined inspection range including a plurality of inspected portions EA (openings 21) by a known phase shift method based on four types of area image data imaged under each of the above pattern lights. Measurement is performed, and the measurement result (area shape data) is stored in the storage unit 57.

In the above series of processes, the inspection range of the metal mask 20 is moved until the area shape data including the shape data related to the number of the inspected portions EA (peripheral portion of the opening 21) required as learning data is acquired. It is repeated while doing it.

When the area shape data including the shape data related to the number of the inspected unit EA required for learning is acquired in step S101, in the following step S102, the learning unit 77 has not learned based on the command from the main control unit 71. Prepare the neural network 90 of. For example, the neural network 90 stored in the storage unit 57 or the like in advance is read out. Alternatively, the neural network 90 is constructed based on the network configuration information (for example, the number of layers of the neural network, the number of nodes of each layer, etc.) stored in the storage unit 57 or the like.

In step S103, learning shape data (three-dimensional shape data for learning) is acquired as learning data. Specifically, based on the command from the main control unit 71, the data processing unit 75 has a large number of inspected units included in the area shape data based on the area shape data stored in the storage unit 57 in step S101. One inspected portion EA is extracted from the EA, and the shape data related to the inspected portion EA is acquired as one learning shape data. Then, this learning shape data is output to the learning unit 77. That is, only the shape data related to the inspected portion EA of the unused metal mask 20 having no abnormality is used as the learning data (learning shape data).

In step S104, the reconstructed shape data is acquired. Specifically, based on the command from the main control unit 71, the learning unit 77 gives the learning shape data acquired in step S103 as input data to the input layer of the neural network 90, whereby the neural network 90 Acquires the reconstructed shape data output from the output layer.

In the following step S105, the learning unit 77 compares the learning shape data acquired in step S103 with the reconstructed shape data output by the neural network 90 in step S104, and whether or not the error is sufficiently small (whether or not the error is sufficiently small). Whether or not it is equal to or less than a predetermined threshold value) is determined.

Here, when the error is sufficiently small, the neural network 90 and its learning information (updated parameters and the like described later) are stored in the storage unit 57 as the AI model 100, and the learning process is terminated.

On the other hand, if the error is not sufficiently small, the network update process (learning of the neural network 90) is performed in step S106, and then the process returns to step S103 again to repeat the above series of processes.

Specifically, in the network update process of step S105, a known learning algorithm such as an error backpropagation method is used so that the loss function representing the difference between the learning shape data and the reconstructed shape data becomes as small as possible. In addition, the weights (parameters) of the filters 94 and 96 in the neural network 90 are updated to more appropriate ones. As the loss function, for example, BCE (Binary Cross-entropy) or the like can be used.

By repeating these processes many times, the neural network 90 makes the error between the learning shape data and the reconstructed shape data as small as possible, and more accurate reconstructed shape data can be output.

Next, the metal mask inspection process performed by the metal mask inspection apparatus 30 will be described with reference to the flowchart of FIG. However, the inspection process shown in FIG. 8 is a process executed for each predetermined inspection range of the metal mask 20.

When the metal mask 20 is carried into the metal mask inspection device 30 and positioned at a predetermined inspection position, the inspection process is started based on the execution of the predetermined inspection program.

When the inspection process is started, first, in step S301, the image data acquisition process is executed. In the present embodiment, in the inspection related to each inspection range on the back surface 201 side of the metal mask 20, the phase of the first pattern light emitted from the first lighting device 32A is changed while the phase of the first pattern light is different from that of the first pattern light. After performing the imaging process four times in, the imaging process is performed four times under the second pattern light having a different phase while changing the phase of the second pattern light emitted from the second lighting device 32B. Acquire 8 types of image data. Hereinafter, it will be described in detail.

As described above, when the metal mask 20 carried into the metal mask inspection device 30 is positioned and fixed at a predetermined inspection position, the movement mechanism control unit 76 first moves the X-axis based on a command from the main control unit 71. The inspection unit 32 is moved by driving and controlling the mechanism 32D and the Y-axis moving mechanism 32E, and the imaging field of view (imaging range) of the camera 32C is adjusted to a predetermined inspection range of the back surface 201 of the metal mask 20.

At the same time, the lighting control unit 72 switches and controls the liquid crystal shutters 32Ab and 32Bb of both lighting devices 32A and 32B, and determines the positions of the first grid and the second grid formed on the liquid crystal shutters 32Ab and 32Bb as a predetermined reference. Set to position.

When the switching setting between the first grid and the second grid is completed, the lighting control unit 72 causes the first light source 32Aa of the first lighting device 32A to emit light, irradiates the first pattern light, and the camera control unit 73 moves the camera. The 32C is driven and controlled to execute the first imaging process under the first pattern light. The image data generated by the image pickup process is taken into the image acquisition unit 74 at any time (the same applies hereinafter). As a result, the area image data of the inspection range including the plurality of inspection portions EA (openings 21) is acquired.

After that, the lighting control unit 72 turns off the first light source 32Aa of the first lighting device 32A and switches the first liquid crystal shutter 32Ab at the same time as the end of the first imaging process under the first pattern light. Run. Specifically, the position of the first grid formed on the first liquid crystal shutter 32Ab is switched from the reference position to the second position where the phase of the first pattern light is shifted by a quarter pitch (90 °). ..

When the switching setting of the first grid is completed, the lighting control unit 72 causes the light source 32Aa of the first lighting device 32A to emit light, irradiates the first pattern light, and the camera control unit 73 drives and controls the camera 32C. , The second imaging process is performed under the first pattern light. After that, by repeating the same process, four types of area image data under the first pattern light having different phases by 90 ° are acquired.

Subsequently, the lighting control unit 72 causes the second light source 32Ba of the second lighting device 32B to emit light and irradiates the second pattern light, and the camera control unit 73 drives and controls the camera 32C to control the second pattern. The first imaging process under light is performed.

After that, the lighting control unit 72 turns off the second light source 32Ba of the second lighting device 32B and switches the second liquid crystal shutter 32Bb at the same time as the end of the first imaging process under the second pattern light. Run. Specifically, the position of the second grid formed on the second liquid crystal shutter 32Bb is switched and set from the reference position to the second position where the phase of the second pattern light is shifted by a quarter pitch (90 °). ..

When the switching setting of the second grid is completed, the lighting control unit 72 causes the light source 32Ba of the second lighting device 32B to emit light, irradiates the second pattern light, and the camera control unit 73 drives and controls the camera 32C. , The second imaging process is performed under the second pattern light. After that, by repeating the same process, four types of area image data under the second pattern light having different phases by 90 ° are acquired.

In the next step S302, the three-dimensional shape data acquisition process is executed. Specifically, based on a command from the main control unit 71, the data processing unit 75 uses a known phase shift method based on four types of area image data imaged under each pattern light in step S301. Three-dimensional shape measurement of a predetermined inspection range including a plurality of inspected portions EA (openings 21) is performed, and the measurement result (area shape data) is stored in the storage unit 57. The shape data acquisition means in the present embodiment is configured by the function of performing such processing. In this embodiment, since the three-dimensional shape measurement is performed by irradiating the pattern light from two directions, it is possible to prevent the occurrence of a shadow portion that is not irradiated with the pattern light.

In the following step S303, the inspection shape data (three-dimensional shape data for inspection) related to each inspected portion EA is acquired.

Specifically, first, based on the command from the main control unit 71, the data processing unit 75 includes a plurality of area shape data included in the area shape data based on the area shape data related to the predetermined inspection range acquired in step S302. All the EA to be inspected are identified and extracted.

Subsequently, these are numbered and registered as the original shape data related to the inspected portion EA. In this process, for example, the original shape data related to the inspected part EA as shown in FIG. 10A and the inspected part EA having some abnormality as shown in FIGS. 10B to 10E. The original shape data and the like will be acquired.

In step S304, the reconstruction process (reconstruction shape data acquisition process) is executed. The function for executing the present process constitutes the reconstructed shape data acquisition means in the present embodiment.

Specifically, based on the command from the main control unit 71, the inspection unit 78 inputs the original shape data related to the inspected unit EA of the predetermined number (for example, 001) acquired in step S303 to the AI model 100. Enter in the layer. Then, the shape data reconstructed by the AI model 100 and output from the output layer is acquired as the reconstructed shape data related to the inspected portion EA of the predetermined number (for example, 001), and the reconstructed shape data is the same. Stored in association with the original shape data of the number. In this way, in this process, the reconstructed shape data is acquired for all the inspected portion EA numbered and registered in step S303.

Here, the AI model 100 is shown in FIGS. 10 (b) to 10 (e), not to mention the case where the original shape data related to the inspected portion EA without any abnormality as shown in FIG. 10 (a) is input. Even when the original shape data related to the inspected part EA with abnormalities is input, as the reconstructed shape data by learning as described above, there is no abnormality as shown in FIG. 10A. The shape data related to the inspection unit EA will be output.

In step S305, a pass / fail determination process is performed. Specifically, based on the command from the main control unit 71, the inspection unit 78 first compares the original shape data having the same number with the reconstructed shape data, and extracts the difference between the two shape data. The function for executing such a process constitutes the comparison means in the present embodiment.

Subsequently, the inspection unit 78 determines whether or not the difference between the two shape data, that is, the portion corresponding to the abnormal portion Q is larger than the predetermined threshold value. Here, when the difference between the two image data is larger than a predetermined threshold value, it is determined that there is an abnormality.

For example, as shown in FIGS. 9A and 9B, when the abnormal portion Q is present in the inspected portion EA on the back surface 201 side of the metal mask 20, the depth d and the width w of the abnormal portion Q are present. When both of the determination items of the above exceed a predetermined threshold value, it is determined that there is an abnormality.

As a result, for example, as shown in FIG. 10B, when the chamfered portion Q1 on the periphery of the opening 21 on the back surface 201 side of the metal mask 20 is excessively rounded due to excessive electrolytic polishing or the like, the chamfered portion is formed. Q1 is also regarded as an abnormal part Q.

Similarly, as shown in FIG. 10 (c), when the inspected portion EA has a scratch Q2 such as a dent or a scrape connected to the opening 21, or as shown in FIG. 10 (d), the opening. Even when the strain Q3 is generated on the peripheral edge of the 21, it is regarded as the abnormal portion Q depending on the degree of the strain Q3.

On the other hand, when the difference between the two image data is smaller than the predetermined threshold value, it is determined that there is no abnormality. The determination means in the present embodiment is configured by the function of performing these determinations.

Here, when the inspection unit 78 determines that there is no abnormality in all the inspected unit EA included in the area shape data (predetermined inspection range of the metal mask 20), the inspection unit 78 determines the inspection range related to the area shape data. Is determined to be "good", and this result is stored in the storage unit 57, and this process is terminated.

On the other hand, when there is at least one inspected part EA determined to be "abnormal" among the plurality of inspected part EA included in the area shape data (predetermined inspection range of the metal mask 20), the said. The inspection range related to the area shape data is determined to be "defective", the result is stored in the storage unit 57, and this process is terminated.

Then, when the metal mask inspection device 30 is determined to be "good" for the entire inspection range as a result of performing the above inspection process for the entire inspection range of the back surface 201 of the metal mask 20, the metal mask 20 has no abnormality. It is determined (pass determination), and the operator is notified to that effect via the display unit 56, the communication unit 58, or the like.

On the other hand, the metal mask inspection apparatus 30 determines that the metal mask 20 has an abnormality when there is at least one inspection range determined to be "defective" among all the inspection ranges on the metal mask 20 (). (Failure determination) is performed, and the operator is notified to that effect via the display unit 56, the communication unit 58, or the like.

As described in detail above, in the present embodiment, of the front and back surfaces of the metal mask 20, the peripheral portion of the opening 21 is from the back surface 201 side, which is the substrate contact surface side that abuts on the printed circuit board 1 during solder printing. The inspection related to the predetermined EA to be inspected can be performed.

As a result, various abnormal parts Q (Q1 to Q4) related to the vicinity of the opening 21 (inspected part EA) on the back surface 201 side of the metal mask 20 which may cause bleeding or sagging of the cream solder 5 during printing are detected. can do. As a result, it is possible to suppress the occurrence of printing defects and the like.

In particular, in the present embodiment, the AI model 100 constructed by learning the neural network 90 is used to determine whether or not there is an abnormal portion Q (Q1 to Q4) in the inspected portion EA of the metal mask 20. This makes it possible to detect minute abnormalities that were difficult to detect in the past.

Further, in the present embodiment, since the original shape data obtained by imaging the inspected portion EA and the reconstructed shape data obtained by reconstructing the original shape data are compared, both are compared. In the shape data, the imaging conditions on the side of the metal mask 20 which is the inspection target (for example, the arrangement position and angle of the metal mask 20, the deflection, etc.) and the imaging conditions on the metal mask inspection device 30 side (for example, the lighting state and the camera 32C). There is no influence based on the difference in the angle of view, etc.), and it becomes possible to detect a finer abnormal portion Q more accurately.

When inspecting a predetermined inspected portion EA (peripheral portion of the opening 21) at a predetermined position of the metal mask 20, the configuration information (position data or position data) of the inspected portion EA is used as a quality determination criterion. In a configuration that requires (dimensional data, etc.), for example, substrate design information such as Gerber data is stored in advance, the configuration information of the inspected part EA to be inspected is appropriately acquired, and the inspection is performed while comparing with the configuration information. Since the quality of the target EA to be inspected is determined, the inspection efficiency may decrease. Further, it is necessary to accurately position the metal mask 20 at the inspection position.

On the other hand, according to the present embodiment, since the AI model 100 learned about the inspected part EA is used to inspect each inspected part EA, each of the large number of inspected part EA is inspected. Since it is not necessary to store individual configuration information in advance and it is not necessary to refer to it at the time of inspection, it is possible to improve inspection efficiency.

In addition, in the present embodiment, the back surface 201, which is the substrate contact surface side of the metal mask 20 that abuts on the printed circuit board 1 during solder printing, is arranged so as to face downward, and is inspected from the lower side of the metal mask 20. The unit 32 is configured to perform the inspection.

As a result, it is less likely to be affected by ambient light such as indoor lighting, and inspection accuracy can be improved. In particular, it is more effective when the inspection is performed in an environment susceptible to minute changes in luminance, such as when the phase shift method is used.

The content is not limited to the description of the above embodiment, and may be implemented as follows, for example. Of course, other application examples and modification examples not illustrated below are also possible.

(A) The configuration of the screen mask is not limited to the metal mask 20 according to the above embodiment, and other configurations may be adopted.

(A-1) For example, the metal mask 20 according to the above embodiment is composed of a mask main body portion 20a and a frame portion 20b, but the present invention is not limited to this, and the frame portion 20b may be omitted.

(A-2) The mask main body 20a according to the above embodiment is made of a metal material (metal). Not limited to this, as the mask main body 20a, a photosensitive emulsion or the like is applied onto a "mesh" formed by weaving threads such as nylon, polyester, and stainless steel to form a covering portion. Even if a type in which the "gauze" is exposed only in the part corresponding to the opening 21 or a composite type in which the metal plate on which the opening 21 is formed is attached on the "gauze (mesh)" is adopted. good.

(B) The configuration related to the production of the printed circuit board 1 such as the production line 10 is not limited to the above embodiment, and other configurations may be adopted.

(B-1) In the above embodiment, the cleaning device 29 and the metal mask inspection device 30 are provided in the vicinity of the solder printing machine 12. Not limited to this, a configuration may be configured in which the functions of the cleaning device and / or the metal mask inspection device are integrally provided with the solder printing machine 12 or the solder printing inspection device 13.

(B-2) The cleaning device 29 and the metal mask inspection device 30 may be provided in a separate room from the production line 10 and the like.

(C) The configuration of the metal mask inspection apparatus 30 is not limited to the above embodiment, and other configurations may be adopted.

For example, in the above embodiment, the inspection unit 32 is arranged from the lower side of the metal mask 20 with the back surface 201, which is the substrate contact surface side of the metal mask 20 that comes into contact with the printed circuit board 1 during solder printing, so as to face downward. It is configured to be inspected by.

Not limited to this, the inspection is performed from the upper side by the inspection unit 32 arranged above the metal mask 20 in a state where the back surface 201 (board contact surface) side of the metal mask 20 is arranged so as to face upward. May be good. However, in order to suppress the influence of ambient light such as indoor lighting and improve the inspection accuracy, it is preferable that the inspection is performed in a state where the back surface 201 side of the metal mask 20 is arranged so as to face downward.

(D) The configuration of the AI model 100 (neural network 90) as the identification means and the learning method thereof are not limited to the above embodiment.

(D-1) Although not particularly mentioned in the above embodiment, when performing the learning process of the neural network 90, the reconstruction process in the metal mask inspection process, or the like, normalization or the like is performed on various data as necessary. It may be configured to perform the processing of.

(D-2) The structure of the neural network 90 is not limited to that shown in FIG. 6, and may be, for example, a configuration in which a pooling layer is provided after the convolution layer 93. Of course, the number of layers of the neural network 90, the number of nodes of each layer, the connection structure of each node, and the like may be different.

(D-3) In the above embodiment, the AI model 100 (neural network 90) is a generative model having a structure of a convolutional autoencoder (CAE), but the present invention is not limited to this, and for example, variational self-coding is performed. It may be a generative model having a structure of a different type of autoencoder such as a VAE (Variational Autoencoder).

(D-4) In the above embodiment, the neural network 90 is learned by the error back propagation method, but the present invention is not limited to this, and various other learning algorithms may be used for learning.

(D-5) The neural network 90 may be configured by an AI processing dedicated circuit such as a so-called AI chip. In that case, only the learning information such as parameters is stored in the storage unit 57, which may be read out by the AI processing dedicated circuit and set in the neural network 90 to configure the AI model 100.

(D-6) In the above embodiment, the learning unit 77 is provided and the neural network 90 is trained in the control device 33, but the present invention is not limited to this, and at least the AI model 100 (learned neural network) is used. The 90) may be stored in the storage unit 57, and the learning unit 77 may be omitted. Therefore, the neural network 90 may be learned outside the control device 33 and stored in the storage unit 57.

(E) The inspection method of the metal mask 20 (inspected portion EA) is not limited to the above embodiment, and other configurations may be adopted.

(E-1) For example, in the above embodiment, whether or not there is an abnormal portion Q in the inspected portion EA by extracting the difference between the original shape data and the reconstructed shape data using the AI model (generated model) 100. It is configured to determine whether or not.

Not limited to this, without using the AI model 100, for example, the reference shape data (configuration information of the inspected part EA such as position data and dimensional data) stored in advance and the calculated shape data of the inspected part EA are compared. By doing so, it may be configured to determine whether or not there is an abnormal portion Q in the inspected portion EA.

(E-2) In the above embodiment, when the abnormal portion Q is present in the inspected portion EA, both the depth d and the width w of the abnormal portion Q exceed a predetermined threshold value, respectively. It is configured to be judged as "abnormal".

Not limited to this, for example, when at least one of the determination items of the depth d and the width w of the abnormality portion Q exceeds a predetermined threshold value, it may be determined that there is an abnormality.

Further, the number of determination items is increased, and when at least one determination item or a plurality of predetermined determination items among the depth d, width w, length l, and volume v of the abnormal portion Q exceeds a predetermined threshold value, " It may be configured to determine "there is an abnormality". Of course, the configuration may be such that only one determination item is set.

(E-3) The type of the abnormal portion Q to be detected is not limited to that exemplified in the above embodiment, and may be configured so that another abnormal portion Q can be detected. For example, in the above embodiment, as the abnormal portion Q, the chamfered portion Q1 [see FIG. 10 (b)] which has been over-polished and becomes too round, the scratch Q2 [see FIG. 10 (c)] connected to the opening 21, and the peripheral edge of the opening 21. Distortion Q3 [see FIG. 10 (d)] and the like are exemplified.

Not limited to this, for example, as shown in FIG. 10 (e), the portion to which the foreign matter Q4 adheres to the inspected portion EA (foreign matter Q4 adhering to the inspected portion EA) may be detected as the abnormal portion Q.

Although it is extremely unlikely that foreign matter Q4 is attached to the unused metal mask 20, for example, after cleaning the metal mask 20 with the cleaning device 29, the metal mask 20 is reattached to the solder printing machine 12. When used, minute residues such as cream solder 5 and flux that could not be removed by cleaning adhered to the peripheral portion (inspected portion EA) of the opening 21 on the back surface 201 side of the metal mask 20. It may remain as it is.

When the foreign matter Q4 adheres to the inspected portion EA in this way, the inspected portion EA (the peripheral portion of the opening 21 on the back surface 201 side of the metal mask 20) floats during solder printing. Cream solder may flow and cause bleeding or sagging.

(E-4) In the above embodiment, when the metal mask 20 (inspected portion EA) is inspected, the pattern light is irradiated to perform three-dimensional measurement. Instead of this, for example, uniform The abnormal portion Q may be detected based on the two-dimensional luminance image data related to the inspected portion EA acquired by imaging while irradiating with light.

At this time, for example, using another AI model (learned neural network) different from the AI model 100, the reflectance (reflected light) differs depending on the uneven shape of the inspected portion EA based on the two-dimensional luminance image data. The shape data related to the part to be inspected EA may be acquired due to the difference in the luminance value) and the like.

(F) The three-dimensional measurement method is not limited to the above embodiment, and other configurations may be adopted.

(F-1) For example, in the above embodiment, in performing three-dimensional measurement by the phase shift method, four types of image data in which the phases of the pattern lights differ by 90 ° are acquired. The number of times and the amount of phase shift are not limited thereto. Other phase shift counts and phase shift amounts that can be measured three-dimensionally by the phase shift method may be adopted.

For example, it may be configured to acquire three types of image data having different phases by 120 ° (or 90 °) and perform three-dimensional measurement, or to acquire two types of image data having different phases by 180 ° (or 90 °). It may be configured to perform three-dimensional measurement.

(F-2) In the above embodiment, the phase shift method is adopted as the three-dimensional measurement method, but the method is not limited to this, and other tertiary methods such as the optical cutting method, the moire method, the focusing method, and the spatial code method are used. The original measurement method may be adopted.

1 ... Printed circuit board, 3 ... Land, 5 ... Cream solder, 10 ... Production line, 12 ... Solder printing machine, 20 ... Metal mask, 21 ... Opening, 29 ... Cleaning device, 30 ... Metal mask inspection device, 32 ... Inspection Unit, 32A ... 1st lighting device, 32B ... 2nd lighting device, 32C ... camera, 33 ... control device, 77 ... learning unit, 78 ... inspection unit, 90 ... neural network, 100 ... AI model, 201 ... back side, 202 ... surface, EA ... inspected part, Q ... abnormal part.

Claims (6)

A screen mask inspection device for inspecting a screen mask that has a plurality of openings and is used when performing solder printing on a substrate.
Of the front and back surfaces of the screen mask, at least a predetermined light can be applied to a predetermined portion to be inspected, which is a peripheral portion of the opening on the back surface side which is the substrate contact surface side which comes into contact with the substrate during solder printing. One irradiation means and
An image pickup means capable of taking an image of the part to be inspected irradiated with the predetermined light, and an image pickup means.
A shape data acquisition means capable of acquiring shape data related to the inspected portion based on one or a plurality of image data of the inspected portion imaged by the image pickup means.
A screen mask inspection apparatus comprising: a determination means capable of determining the quality of the inspected portion based on the shape data related to the inspected portion acquired by the shape data acquisition means. The determination means is
The screen according to claim 1, wherein the quality of the inspected portion can be determined by determining whether or not the size of the chamfered portion on the peripheral edge of the opening exceeds a predetermined amount. Mask inspection device. The determination means is
Claim 1 or 2 is characterized in that the quality of the inspected portion can be determined by determining whether or not the inspected portion has an abnormal portion exceeding a predetermined size. The screen mask inspection device described. For a neural network having a coding unit that extracts a feature amount from the input shape data and a decoding unit that reconstructs the shape data from the feature amount, only the shape data related to the inspected part having no abnormality is learned. Equipped with identification means generated by training as data,
The determination means is
A reconstructed shape capable of acquiring reconstructed shape data related to the inspected portion reconstructed by inputting the shape data related to the inspected portion acquired by the shape data acquisition means into the identification means as original shape data. Data acquisition method and
A comparison means capable of comparing the original shape data and the reconstructed shape data is provided.
The screen mask inspection apparatus according to any one of claims 1 to 3, wherein the quality of the inspected portion can be determined based on the comparison result by the comparison means. The irradiation means is configured to be capable of irradiating light for three-dimensional measurement as the predetermined light.
The shape data acquisition means according to any one of claims 1 to 4, wherein the shape data acquisition means is configured to be able to acquire the three-dimensional shape data related to the inspected portion by using a predetermined three-dimensional measurement method. The screen mask inspection device described. The screen mask is arranged so that the back surface side is the lower side.
The irradiation means is arranged so as to be able to irradiate the predetermined light from below with respect to the back surface side of the screen mask.
The screen mask inspection apparatus according to any one of claims 1 to 5, wherein the image pickup means is arranged so that the back surface side of the screen mask can be imaged from below.

Images