JP2022050122A - Tablet printing device and tablet printing method - Google Patents

Abstract

To provide a tablet printing device and a tablet printing method which accurately determine the shape, position or rotation angle of a tablet and thereby accurately perform printing on the surface of the tablet.SOLUTION: A tablet printing device 1 includes an auto encoder 71. The auto encoder 71 outputs a reconstruction image D3 by performing dimension reduction and reconstruction on a photographed image D1 of a tablet. A parameter used for processing of the dimension reduction and reconstruction is adjusted with machine learning in advance. The reconstruction image D3 output from the auto encoder 71 is the image in which random elements such as glossiness, roughness, defects and the like are reduced. Therefore, the shape, position or rotation angle of the tablet can be accurately determined on the basis of the reconstruction image D3. Consequently, printing can be accurately performed on the surface of the tablet.SELECTED DRAWING: Figure 5

Description

The present invention relates to a tablet printing apparatus and a tablet printing method for printing on the surface of a tablet while transporting the tablet.

Conventionally, a tablet printing device for printing images such as characters and logo marks on the surface of tablets such as pharmaceuticals and supplements by an inkjet method has been known. A conventional tablet printing apparatus is described in, for example, Patent Document 1.

In the tablet printing device, each tablet is photographed while transporting a plurality of tablets in an aligned state. Then, based on the obtained captured image, the shape, position, rotation angle, etc. are determined for each tablet. Then, according to the above determination result, printing is performed on the tablet while finely adjusting the position of the ink droplet to be ejected on each tablet. For example, in the apparatus of Patent Document 1, the rotation angle of the score line of the tablet is determined based on the photographed image of the tablet, and printing is performed along the score line.

Japanese Unexamined Patent Publication No. 2017-200495

However, the shape and surface condition of a plurality of tablets transported in this type of tablet printing device are not strictly constant. Therefore, in the photographed image of the tablet, irregular gloss or roughness due to the roughness of the drug component may appear on the surface of the tablet. In addition, some tablets may have defects such as chips and black spots. Therefore, due to the influence of these gloss, roughness, or defects, the shape, position, or rotation angle of the tablet may not be accurately determined.

The present invention has been made in view of such circumstances, and can accurately determine the shape, position, or rotation angle of a tablet, thereby printing on the surface of the tablet with high accuracy. It is an object of the present invention to provide an apparatus and a tablet printing method.

In order to solve the above problems, the first invention of the present application is a tablet printing device that prints on the surface of a tablet while transporting the tablet, wherein the transport mechanism for transporting the tablet and the tablet transported by the transport mechanism are used. An imaging unit that shoots and acquires a shot image, an auto-encoder that outputs a reconstructed image by reducing dimensions and reconstructing the shot image, and a tablet based on the reconstructed image. The auto encoder is provided with a determination unit for determining at least one of the shape, position, and rotation angle of the tablet, and a printing unit for printing on the surface of the tablet based on the determination result of the determination unit. The dimension reduction and the reconstruction are performed using parameters adjusted in advance by machine learning so that the difference between the image of the tablet and the image of the output tablet is small.

The second invention of the present invention is the tablet printing apparatus of the first invention, and the determination unit determines the rotation angle of a tablet based on the score line of the tablet included in the reconstructed image.

The third aspect of the present invention is the tablet printing apparatus of the first invention or the second invention, wherein an inspection area setting unit for setting an inspection area of a tablet based on the reconstructed image, and the inspection of the photographed image. In the inspection area set by the area setting unit, an inspection unit for inspecting the defect of the tablet is further provided.

The fourth aspect of the present invention is the tablet printing apparatus of the third invention, in which the inspection area setting unit sets a plurality of the inspection areas based on the reconstructed image, and the inspection unit sets each inspection area. Inspect tablets for defects under different conditions.

A fifth aspect of the present invention is the tablet printing apparatus of the fourth invention, wherein the plurality of inspection areas include a region corresponding to an edge of a tablet.

The sixth invention of the present application is a tablet printing method in which printing is performed on the surface of a tablet while transporting the tablet, and a) an image of the tablet is input to an auto encoder that performs dimension reduction and reconstruction. The step of adjusting the parameters of the auto encoder so that the difference between the image of the tablet and the reconstructed image of the output tablet is small, and b) the tablet is photographed while the tablet is being conveyed, and the photographed image is acquired. And c) the step of inputting the photographed image into the auto-encoder whose parameters have been adjusted in the step a) and obtaining the reconstructed image output from the auto-encoder, and d) the reconstructed image. Based on the above, a step of determining at least one of the shape, position, and rotation angle of the tablet, and e) a step of printing on the surface of the tablet based on the determination result of the step d) are provided.

The seventh invention of the present application is the tablet printing method of the sixth invention, and in the step d), the rotation angle of a tablet is determined based on the score line of the tablet contained in the reconstructed image.

The eighth invention of the present application is the tablet printing method of the sixth invention or the seventh invention, in which f) a step of setting an inspection area of a tablet based on the reconstructed image, and g) the photographed image, said. In the inspection area set by the step f), a step of inspecting the defect of the tablet is further provided.

The ninth aspect of the present invention is the tablet printing method of the eighth invention, in which a plurality of the inspection areas are set based on the reconstructed image in the step f), and for each inspection area in the step g). Inspect tablets for defects under different conditions.

The tenth invention of the present application is the tablet printing method of the ninth invention, and the plurality of inspection areas include the area corresponding to the edge of a tablet.

According to the first to tenth inventions of the present application, the autoencoder can generate a reconstructed image in which random elements such as gloss, roughness, and defects are reduced from the captured image. Therefore, the shape, position, or rotation angle of the tablet can be accurately determined based on the reconstructed image. Therefore, printing can be performed accurately on the surface of the tablet.

In particular, according to the third or eighth invention of the present application, the inspection area can be set accurately based on the reconstructed image. Therefore, the defect of the tablet can be inspected with high accuracy.

In particular, according to the fourth or ninth invention of the present application, a plurality of inspection areas can be set accurately based on the reconstructed image.

In particular, according to the fifth or tenth invention of the present application, the inspection region corresponding to the edge of the tablet where the deviation is likely to occur can be set accurately by using the reconstructed image.

It is a figure which showed the structure of the tablet printing apparatus. It is a partial perspective view of a transport mechanism. It is a bottom view of a head. It is a block diagram which showed the electrical connection between a computer and each part of a tablet printing apparatus. It is a block diagram which conceptually showed the function of a computer. It is a figure which conceptually showed the dimension reduction and reconstruction processing performed by an autoencoder. It is a figure which showed the example of the inspection area. It is a flowchart which shows the flow of machine learning. It is a flowchart which shows the flow of a print process.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the following description, the direction in which a plurality of tablets are transported is referred to as a "transport direction", and the direction perpendicular to and horizontal to the transport direction is referred to as a "width direction".

<1. Overall configuration of printing equipment>
FIG. 1 is a diagram showing a configuration of a tablet printing apparatus 1 according to an embodiment of the present invention. The tablet printing device 1 is a device that prints images such as a product name, a product code, a company name, and a logo mark on the surface of each tablet 9 while transporting a plurality of tablets 9. As shown in FIG. 1, the tablet printing apparatus 1 of the present embodiment includes a transport mechanism 10, a first imaging unit 20, a printing unit 30, a second imaging unit 40, a defective product discharging unit 50, a non-defective product discharging unit 60, and a computer. It is equipped with 70.

The transport mechanism 10 is a mechanism for transporting a plurality of tablets 9 while holding them in an aligned state. The transport mechanism 10 has a pair of pulleys 11 and an annular transport belt 12 spanned between the pair of pulleys 11. The plurality of tablets 9 charged into the tablet printing device 1 are arranged at equal intervals by a carry-in mechanism 80 composed of a vibration feeder, a transfer drum, and the like, and are supplied to the surface of the transfer belt 12. One of the pair of pulleys 11 is rotated by the power obtained from the transport motor 13. As a result, the transport belt 12 rotates in the direction of the arrow in FIG. At this time, the other of the pair of pulleys 11 is driven to rotate with the rotation of the transport belt 12.

FIG. 2 is a partial perspective view of the transport mechanism 10. As shown in FIG. 2, the transport belt 12 is provided with a plurality of suction holes 14. The plurality of suction holes 14 are arranged at equal intervals in the transport direction and the width direction. Further, as shown in FIG. 1, the transport mechanism 10 has a suction mechanism 15 that sucks gas from the space inside the transport belt 12. When the suction mechanism 15 is operated, the space inside the transport belt 12 becomes a negative pressure lower than the atmospheric pressure. The plurality of tablets 9 are adsorbed and held in the adsorption holes 14 by this negative pressure.

In this way, the plurality of tablets 9 are held on the surface of the transport belt 12 in a state of being aligned in the transport direction and the width direction. Then, the transport mechanism 10 transports the plurality of tablets 9 in the transport direction by rotating the transport belt 12. Below the four heads 31, which will be described later, the plurality of tablets 9 are horizontally conveyed.

The first imaging unit 20 is a portion for acquiring a photographed image of the tablet 9 before printing. The first imaging unit 20 holds a plurality of tablets 9 transported by the transport belt 12 at the first image pickup position P1 on the downstream side of the transport path from the carry-in mechanism 80 and on the upstream side of the transport path from the print position P2 described later. , Take a picture from above. For the first image pickup unit 20, for example, a line sensor in which image pickup elements such as CCD and CMOS are arranged in the width direction is used. However, another image pickup device such as an area sensor may be used for the first image pickup unit 20. The captured image acquired by photographing is transmitted from the first imaging unit 20 to the computer 70 described later. The computer 70 has no defects such as the presence / absence of the tablet 9 in each suction hole 14, the position of the tablet 9, the rotation angle of the tablet 9, and the chip of the tablet 9 based on the photographed image obtained from the first imaging unit 20. Inspect whether or not.

The printing unit 30 prints an image on the surface of the tablet 9 by an inkjet method at the printing position P2 on the downstream side of the transport path from the first image pickup position P1 and on the upstream side of the transport path from the second image pickup position P3 described later. It is a processing unit. As shown in FIG. 1, the printing unit 30 of this embodiment has four heads 31. The four heads 31 are located above the transport belt 12 and are arranged in a row along the transport direction of the tablets 9. The four heads 31 eject ink droplets of different colors (eg, cyan, magenta, yellow, and black) toward the surface of the tablet 9. Then, a multicolor image is printed on the surface of the tablet 9 by superimposing the monochromatic images formed by each of these colors. As the ink discharged from each head 31, edible ink manufactured from raw materials approved by the Japanese Pharmacopoeia, the Food Sanitation Law, etc. is used.

FIG. 3 is a bottom view of one head 31. In FIG. 3, the transport belt 12 and the plurality of tablets 9 held by the transport belt 12 are shown by a two-dot chain line. As shown enlarged in FIG. 3, a plurality of nozzles 311 capable of ejecting ink droplets are provided on the ejection surface 310 which is the lower surface of the head 31. In the present embodiment, a plurality of nozzles 311 are two-dimensionally arranged on the lower surface of the head 31 in the transport direction and the width direction. The nozzles 311 are arranged so as to be displaced in the width direction. By arranging the plurality of nozzles 311 two-dimensionally in this way, the positions of the nozzles 311 in the width direction can be brought close to each other. However, the plurality of nozzles 311 may be arranged in a row along the width direction.

As the method of ejecting ink droplets from the nozzle 311, for example, a so-called piezo method is used in which the ink in the nozzle 311 is pressurized and ejected by applying a voltage to the piezo element to deform it. However, the ink droplet ejection method may be a so-called thermal system in which the heater is energized and the ink in the nozzle 311 is heated and expanded to eject the ink droplets.

The second image pickup unit 40 is a portion for acquiring a photographed image of the tablet 9 after printing. The second imaging unit 40 is a plurality of tablets transported by the transport belt 12 at the second image pickup position P3 on the downstream side of the transport path from the print position P2 and on the upstream side of the transport path from the defective product discharge position P4 described later. 9 is photographed from above. For the image pickup unit 40, for example, a line sensor in which image pickup elements such as CCD and CMOS are arranged in the width direction is used. However, another image pickup device such as an area sensor may be used for the second image pickup unit 40. The captured image acquired by photographing is transmitted from the second imaging unit 40 to the computer 70 described later. The computer 70 inspects the quality of the printed image formed on the surface of the tablet 9 based on the second photographed image obtained from the second image pickup unit 40.

The defective product discharge unit 50 disposes of the tablet 9 determined to be defective at the defective product discharge position P4 on the downstream side of the transport path from the second imaging position P3 and on the upstream side of the transport path from the non-defective product discharge position P5 described later. This is the part to be discharged. The defective product discharging unit 50 has a defective product blow mechanism 51 arranged inside the transport belt 12, and a defective product recovery unit 52 arranged below the defective product blow mechanism 51 with the transport belt 12 sandwiched between them. The defective product blow mechanism 51 blows gas only to the suction holes 14 holding the tablets 9 determined to be defective products among the plurality of suction holes 14 of the transport belt 12. Then, the suction hole 14 has a positive pressure higher than the atmospheric pressure. As a result, the adsorption of the tablet 9 in the suction hole 14 is released, and the tablet 9 determined to be a defective product falls from the transport belt 12 to the defective product collection unit 52.

The non-defective product discharge unit 60 is a portion for discharging the tablet 9 determined to be a non-defective product at the non-defective product discharge position P5 on the downstream side of the transport path from the defective product discharge position P4. The non-defective product discharging unit 60 has a non-defective product blow mechanism 61 arranged inside the transport belt 12, and a non-defective product collection unit 62 arranged below the non-defective product blow mechanism 61 with the transport belt 12 sandwiched between them. The non-defective blow mechanism 61 blows gas toward the plurality of suction holes 14 of the transport belt 12. Then, the suction hole 14 has a positive pressure higher than the atmospheric pressure. As a result, the adsorption of the tablet 9 in the suction hole 14 is released, and the tablet 9 determined to be a non-defective product is dropped from the transport belt 12 to the non-defective product collection unit 62.

The computer 70 is an information processing device for controlling the operation of each part in the tablet printing device 1 and performing machine learning described later. FIG. 4 is a block diagram showing an electrical connection between the computer 70 and each part of the tablet printing apparatus 1. As conceptually shown in FIG. 4, the computer 70 has a processor 701 such as a CPU, a memory 702 such as a RAM, and a storage unit 703 such as a hard disk drive. A computer program CP for executing a print process and machine learning, which will be described later, is stored in the storage unit 703. Further, the trained model M obtained by machine learning is also stored in the storage unit 703.

The computer 70 can communicate with the above-mentioned transport mechanism 10, the first imaging unit 20, four heads 31, the second imaging unit 40, the defective blow mechanism 51, the non-defective blow mechanism 61, and the carry-in mechanism 80 by wire or wirelessly. It is connected to the. The computer 70 controls the operation of each of these parts according to the computer program CP and the trained model M.

<2. About computers>
FIG. 5 is a block diagram conceptually showing the functions of the computer 70 described above. As shown in FIG. 5, the computer 70 includes an autoencoder 71, a determination unit 72, a print instruction unit 73, an inspection area setting unit 74, and an inspection unit 75. The functions of the autoencoder 71, the determination unit 72, the print instruction unit 73, the inspection area setting unit 74, and the inspection unit 75 are realized by operating the processor 701 of the computer 70 according to the computer program CP.

The autoencoder 71 is a processing unit that outputs the reconstructed image D3 by performing dimension reduction (encoding) and reconstruction (decoding) of the input captured image D1 by a multi-layer neural network. FIG. 6 is a diagram conceptually showing the dimension reduction and reconstruction processing executed by the autoencoder 71.

As shown in FIG. 6, the dimension reduction is a process of compressing the input high-dimensional captured image D1 into low-dimensional data called a latent variable D2. The reconstruction is a process of reproducing the image of the tablet 9 from the latent variable D2 to generate the reconstructed image D3. The autoencoder 71 reduces random information such as gloss, roughness, and defects while retaining essential information such as the shape, color, and orientation of the score line of the tablet 9 in the dimension reduction process. Therefore, the reconstructed image D3 generated by the reconstructed image is an image of the tablet 9 in which random elements such as gloss, roughness, and defects are reduced.

The autoencoder 71 can adjust the parameters used for the dimension reduction and reconstruction processes by performing machine learning. In machine learning, a large number of good tablet 9 captured images D1 are used, and the difference between the captured image D1 of the tablet 9 input to the autoencoder 71 and the reconstructed image D3 of the tablet 9 output from the autoencoder 71 is Adjust the dimensionality reduction and reconstruction parameters to be smaller. This gives a trained model M containing the tuned parameters.

The trained model M is stored in the storage unit 703 of the computer 70. When printing the tablet 9 in the tablet printing apparatus 1, the autoencoder 71 performs the dimension reduction and reconstruction processing using the parameters indicated by the trained model M obtained by prior machine learning. Therefore, when the photographed image D1 of the tablet 9 acquired by the first imaging unit 20 is input to the autoencoder 71, the reconstructed image D3 of the tablet 9 in which random elements such as gloss, roughness, and defects are reduced is input to the autoencoder 71. Can be output from.

The determination unit 72 is a processing unit that determines the shape, position, and rotation angle of the tablet 9 transported by the transport mechanism 10. The determination unit 72 determines the shape, position, and rotation angle of the tablet 9 by performing image processing on the reconstructed image D3 output from the autoencoder 71. Specifically, the determination unit 72 extracts a portion corresponding to the edge of the tablet 9 from the reconstructed image D3, and determines the shape and position of the tablet 9 based on the shape and position of the edge. Further, the determination unit 72 extracts a portion corresponding to the score line of the tablet 9 from the reconstructed image D3, and determines the rotation angle of the tablet 9 based on the shape and position of the score line. In the present application, the "rotation angle" means the rotation angle of the tablet 9 with respect to an axis that passes through the center of the tablet 9 and is perpendicular to the transport direction and the width direction. The determination unit 72 outputs these determination results D4 to the print instruction unit 73.

The print instruction unit 73 is a processing unit for causing the print unit 30 to execute the ink ejection process. The determination result D4 of the determination unit 72 and the image data D5 indicating the image to be printed on the tablet 9 are input to the print instruction unit 73. The print instruction unit 73 adjusts the image data D5 according to the determination result D4 input from the determination unit 72. Specifically, the print instruction unit 73 shifts the coordinates of the image data D5 according to the position of the tablet 9 indicated by the determination result D4. Further, the print instruction unit 73 rotates the image data D5 according to the rotation angle of the tablet 9 indicated by the determination result D4. Then, the print instruction unit 73 outputs the print instruction D6 to the plurality of heads 31 of the print unit 30 based on the adjusted image data D5.

The plurality of heads 31 eject ink droplets on the upper surface of the tablet 9 to be conveyed based on the print instruction D6 output from the print instruction unit 73. As a result, the image is printed at an appropriate position on the surface of the tablet 9 at an appropriate rotation angle.

The storage unit 703 of the computer 70 may store a plurality of image data D5 having different rotation angles in advance. In that case, the print instruction unit 73 may select and read the image data D5 having an appropriate rotation angle from the storage unit 703 according to the rotation angle of the tablet 9 indicated by the determination result D4. By doing so, the processing load of the computer 70 related to the rotation of the image data D5 can be reduced.

The inspection area setting unit 74 is a processing unit for setting an area to be inspected (hereinafter referred to as “inspection area”) of the captured image D1. FIG. 7 is a diagram showing an example of an inspection area of the captured image D1. In the example of FIG. 7, with respect to the photographed image D1, the inspection area A1 corresponding to the edge of the tablet 9, the two semicircular inspection areas A2 and A3 surrounded by the score line and the edge of the tablet 9, and the tablet. An inspection area A4 corresponding to the side surface of 9 is set. The inspection area setting unit 74 specifies the coordinates of the portion corresponding to the edge of the tablet 9 and the portion corresponding to the score line in the reconstructed image D3 output from the autoencoder 71 by image processing. Then, a plurality of inspection areas A1 to A4 in the captured image D1 are set based on the specified coordinates. That is, the image to be inspected is the captured image D1, but the inspection area setting unit 74 sets the inspection areas A1 to A4 in the captured image D1 by using the reconstructed image D3.

The inspection unit 75 is a processing unit that inspects the defect of the tablet 9 based on the photographed image D1. The photographed image D1 to be inspected may be an image of the tablet 9 before printing acquired by the first imaging unit 20, or an image of the tablet 9 after printing acquired by the second imaging unit 40. May be good. The inspection unit 75 inspects the defect of the tablet 9 in each of the plurality of inspection regions A1 to A4 set by the inspection region setting unit 74 described above in the captured image D1.

As an inspection method, for example, a photographed image D1 is compared with a reference image which is an image of a normal tablet 9, and a portion where the difference between the pixel values of both images is larger than a preset threshold value is detected as a defect. do. Alternatively, the captured image D1 and the reconstructed image D3 output from the autoencoder 71 may be compared, and a portion where the difference between the pixel values of the two images is larger than the preset threshold value may be detected as a defect. ..

In the present embodiment, a plurality of inspection areas A1 to A4 are set for one captured image D1. Therefore, the defect of the tablet 9 can be inspected under different conditions for each of the inspection areas A1 to A4. For example, the threshold value can be relaxed for an inspection area where erroneous detection is likely to occur. Further, the threshold value can be made strict for the inspection area where erroneous detection is unlikely to occur. As a result, accurate inspection can be performed while suppressing erroneous detection of defects. However, the number of inspection areas set by the inspection area setting unit 74 may be one for one captured image D1.

<3. About machine learning ＞
Subsequently, the machine learning performed in the autoencoder 71 will be described. FIG. 8 is a flowchart showing the flow of machine learning. This machine learning is executed in advance before the printing process for the plurality of tablets 9 is executed in the tablet printing apparatus 1.

When performing machine learning, a large number (for example, 100 to 1000) of photographed images D1 of non-defective non-defective tablets 9 are prepared. Then, as shown in FIG. 8, the photographed image D1 of the non-defective tablet 9 is input to the autoencoder 71 (step S1). The autoencoder 71 generates a reconstructed image D3 by executing a dimension reduction and reconstructing process on the input captured image D1 (step S2). When the reconstructed image D3 is output from the autoencoder 71, the computer 70 processes each of dimension reduction and reconstruction so that the difference between the captured image D1 and the reconstructed image D3 becomes small based on the machine learning algorithm. The parameters of (step S3) are adjusted.

The computer 70 repeats the above steps S1 to S3 until a predetermined end condition is satisfied (step S4: No). The end condition may be, for example, that the difference between the captured image D1 and the reconstructed image D3 converges within a preset allowable range. Further, the end condition may be that the number of times the learning process of steps S1 to S3 is repeatedly executed reaches a preset number of times.

Eventually, when the end condition is satisfied, the computer 70 ends machine learning (step S4: Yes). This gives a trained model M containing the tuned parameters. The autoencoder 71 obtained from the trained model M can convert the photographed image D1 of the tablet 9 into a reconstructed image D3 of the tablet 9 in which random elements such as gloss, roughness, and defects are reduced and output. It becomes.

<4. About printing process>
Subsequently, the printing process executed in the tablet printing apparatus 1 after the completion of the machine learning described above will be described. FIG. 9 is a flowchart showing the flow of the printing process. The tablet printing device 1 performs processing while sequentially transporting a plurality of tablets 9. However, FIG. 9 shows the flow of processing when one tablet 9 is focused on.

At the time of executing the printing process, first, the tablet 9 is carried onto the upper surface of the transport belt 12 via the carrying-in mechanism 80 (step S5). Then, the tablet 9 is conveyed toward the downstream side in the conveying direction by the rotation of the conveying belt 12. When the tablet 9 reaches the first imaging position P1 on the transport path, the first imaging unit 20 photographs the tablet 9. As a result, the first imaging unit 20 acquires the photographed image D1 of the tablet 9 before printing (step S6). The first image pickup unit 20 transmits the obtained captured image D1 to the computer 70.

The computer 70 inputs the captured image D1 obtained from the first imaging unit 20 to the autoencoder 71. The autoencoder 71 performs dimension reduction and reconstruction processing on the input captured image D1 using the parameters indicated by the trained model M. As a result, a reconstructed image D3 of the tablet 9 in which random elements such as gloss, roughness, and defects are reduced is generated (step S7). The autoencoder 71 outputs the generated reconstructed image D3 to the determination unit 72 and the inspection area setting unit 74.

Subsequently, the determination unit 72 determines the shape, position, and rotation angle of the tablet 9 by performing image processing on the reconstructed image D3 (step S8). Specifically, the determination unit 72 extracts a portion corresponding to the edge of the tablet 9 from the reconstructed image D3, and determines the shape and position of the tablet 9 based on the shape and position of the edge. Further, the determination unit 72 extracts a portion corresponding to the score line of the tablet 9 from the reconstructed image D3, and determines the rotation angle of the tablet 9 based on the shape and position of the score line. Then, the determination unit 72 outputs these determination results D4 to the print instruction unit 73.

In this way, the determination unit 72 determines the shape, position, and rotation angle of the tablet 9 based on the reconstructed image D3 instead of the captured image D1. Therefore, the shape, position, and rotation angle of the tablet 9 can be accurately determined by suppressing the influence of random elements such as gloss, roughness, and defects.

Subsequently, the print instruction unit 73 adjusts the image data D5 according to the determination result D4 input from the determination unit 72 (step S9). Specifically, the print instruction unit 73 shifts the coordinates of the image data D5 according to the position of the tablet 9 indicated by the determination result D4. Further, the print instruction unit 73 rotates the image data D5 according to the rotation angle of the tablet 9 indicated by the determination result D4. Then, the print instruction unit 73 outputs the print instruction D6 to the plurality of heads 31 of the print unit 30 based on the adjusted image data D5.

When the tablet 9 reaches the print position P2 on the transport path, the plurality of heads 31 eject ink droplets on the upper surface of the tablet 9 according to the print instruction D6 output from the print instruction unit 73. Thereby, the image shown by the image data D5 is printed at an appropriate position on the surface of the tablet 9 at an appropriate rotation angle (step S10).

After that, when the tablet 9 reaches the second imaging position P3 on the transport path, the second imaging unit 40 photographs the tablet 9. As a result, the second imaging unit 40 acquires the photographed image D1 of the tablet 9 after printing (step S11). The second image pickup unit 40 transmits the obtained captured image D1 to the computer 70.

Subsequently, the inspection area setting unit 74 of the computer 70 sets the inspection areas A1 to A4 for the captured image D1 (step S12). Here, the photographed image D1 to be inspected is both an image of the tablet 9 before printing obtained from the first imaging unit 20 and an image of the tablet 9 after printing obtained from the second imaging unit 40. The inspection area setting unit 74 sets a plurality of inspection areas A1 to A4 based on the reconstructed image D3 output from the autoencoder 71. Therefore, it is possible to accurately set a plurality of inspection areas A1 to A4 by suppressing the influence of random elements such as gloss, roughness, and defects. In particular, if the inspection region A1 corresponding to the edge of the tablet 9 is set based on the photographed image D1 itself, it is likely to be displaced due to the influence of the gloss, roughness, chipping, etc. of the tablet 9. However, if the reconstructed image D3 is used, the inspection region A1 can be set accurately by suppressing the influence of the gloss, roughness, or chipping of the tablet 9.

After that, the inspection unit 75 of the computer 70 inspects the defect of the tablet 9 in each of the plurality of inspection regions A1 to A4 set in the above-mentioned step S12 in the captured image D1 (step S13). Thereby, it is determined whether the tablet 9 is a non-defective product without defects or a defective product with defects. As described above, the plurality of inspection areas A1 to A4 are accurately set based on the reconstructed image D3. Therefore, the inspection unit 75 can accurately inspect the defect of the tablet 9 in each region. The inspection unit 75 may inspect the defect of the tablet 9 under different conditions for each of the inspection areas A1 to A4.

When the tablet 9 is determined to be a defective product, the computer 70 operates the defective product blowing mechanism 51 when the tablet 9 reaches the defective product discharge position P4 on the transport path. The defective product blow mechanism 51 blows gas to the suction holes 14 holding the tablets 9 determined to be defective products among the plurality of suction holes 14 of the transport belt 12. As a result, the tablet 9 determined to be defective is discharged from the transport belt 12 to the defective product collection unit 52 (step S14).

When the tablet 9 is determined to be a non-defective product, the computer 70 operates the non-defective product blow mechanism 61 when the tablet 9 reaches the non-defective product discharge position P5 on the transport path. The non-defective blow mechanism 61 blows gas toward the plurality of suction holes 14 of the transport belt 12. As a result, the tablet 9 determined to be a non-defective product is discharged from the transport belt 12 to the non-defective product collection unit 62 (step S15).

As described above, in the tablet printing apparatus 1, the autoencoder 71 can generate a reconstructed image D3 in which random elements such as gloss, roughness, and defects are reduced from the captured image D1. Then, the generated reconstructed image D3 is used to determine the shape, position, and rotation angle of the tablet 9. Therefore, the shape, position, and rotation angle of the tablet 9 can be accurately determined by suppressing the influence of the above random elements. Therefore, printing can be performed accurately on the surface of the tablet 9.

Further, in the tablet printing apparatus 1, the reconstructed image D3 generated by the autoencoder 71 is used for setting the inspection areas A1 to A4. Therefore, the inspection areas A1 to A4 can be set accurately while suppressing the influence of the above random elements. Therefore, the defect of the tablet 9 can be inspected with high accuracy.

<5. Modification example>
Although the embodiment of the present invention has been described above, the present invention is not limited to the above embodiment.

In the above embodiment, the determination unit 72 of the computer 70 determines the shape, position, and rotation angle of the tablet 9. However, the determination unit 72 does not necessarily have to determine all of the shape, position, and rotation angle of the tablet 9. The determination unit 72 may determine at least one of the shape, position, and rotation angle of the tablet 9 for the printing process by the printing unit 30.

Further, in the above embodiment, the inspection area setting unit 74 of the computer 70 sets a plurality of inspection areas A1 to A4 for the captured image D1. Then, the inspection unit 75 inspected each of these inspection areas A1 to A4. However, there may be a region not to be inspected in the plurality of inspection regions A1 to A4.

Further, the inspection method by the inspection unit 75 is not limited to the comparative inspection exemplified in the above embodiment.

Further, the autoencoder 71 is not limited to a simple autoencoder, but may be an advanced autoencoder such as VAE (Variational Auto Encoder) or UR (UnRegularized anomaly score) -VAE.

Further, the tablet printing device 1 may have a drying mechanism for drying the ink adhering to the tablet 9, in addition to the configuration of the above embodiment. The drying mechanism may be provided on the downstream side in the transport direction from the printing unit 30 and on the upstream side in the transport direction from the defective product discharge unit 50. The drying mechanism may be, for example, a mechanism for blowing hot air toward the tablets 9 transported by the transport belt 12.

Further, in the above embodiment, the printing unit 30 is provided with four heads 31. However, the number of heads 31 included in the printing unit 30 may be 1 to 3, or 5 or more.

Further, the "tablet" in the present invention may be not only a tablet as a pharmaceutical product but also a tablet as a supplement or a tablet confectionery such as ramune.

In addition, the details of the tablet printing apparatus and the tablet printing method can be appropriately changed without departing from the spirit of the present invention. Further, the elements appearing in the above-described embodiments and modifications may be appropriately combined as long as there is no contradiction.

1 Tablet printing device 9 Tablets 10 Transport mechanism 20 1st image pickup unit 30 Printing section 31 Head 40 2nd image pickup section 50 Defective product ejection unit 60 Good product ejection unit 70 Computer 71 Autoencoder 72 Judgment unit 73 Printing instruction unit 74 Inspection area setting unit 75 Inspection unit 80 Carry-in mechanism A1 to A4 Inspection area D1 Captured image D2 Latent variable D3 Reconstructed image D4 Judgment result D5 Image data D6 Print instruction M Trained model

Claims (10)

A tablet printing device that prints on the surface of tablets while transporting them.
A transport mechanism for transporting tablets and
An image pickup unit that photographs a tablet transported by the transport mechanism and acquires a photographed image.
An autoencoder that outputs a reconstructed image by reducing the dimensions and reconstructing the captured image.
A determination unit that determines at least one of the shape, position, and rotation angle of the tablet based on the reconstructed image.
A printing unit that prints on the surface of the tablet based on the determination result of the determination unit,
Equipped with
The autoencoder performs the dimensionality reduction and the reconstruction using parameters pre-adjusted by machine learning so that the difference between the image of the input tablet and the image of the output tablet is small. Tablet printing device. The tablet printing apparatus according to claim 1.
The determination unit is a tablet printing device that determines the rotation angle of a tablet based on the score line of the tablet included in the reconstructed image. The tablet printing apparatus according to claim 1 or 2.
An inspection area setting unit that sets an inspection area for tablets based on the reconstructed image,
In the inspection area set by the inspection area setting unit of the photographed image, an inspection unit for inspecting a tablet defect and an inspection unit.
Further equipped with a tablet printing device. The tablet printing apparatus according to claim 3.
The inspection area setting unit sets a plurality of the inspection areas based on the reconstructed image, and sets the inspection area.
The inspection unit is a tablet printing device that inspects for defects in tablets under different conditions for each inspection area. The tablet printing apparatus according to claim 4.
The tablet printing apparatus, wherein the plurality of inspection areas include a region corresponding to an edge of a tablet. It is a tablet printing method that prints on the surface of a tablet while transporting the tablet.
a) A tablet image is input to the autoencoder that performs dimension reduction and reconstruction, and the parameters of the autoencoder are such that the difference between the input tablet image and the output tablet reconstruction image is small. And the process of adjusting
b) The process of photographing the tablet while transporting the tablet and acquiring the photographed image,
c) In the step a), the captured image is input to the autoencoder whose parameters have been adjusted, and a reconstructed image output from the autoencoder is obtained.
d) A step of determining at least one of the shape, position, and rotation angle of the tablet based on the reconstructed image.
e) A step of printing on the surface of the tablet based on the determination result of the step d), and
A tablet printing method. The tablet printing method according to claim 6.
In the step d), a tablet printing method for determining the rotation angle of a tablet based on the score line of the tablet contained in the reconstructed image. The tablet printing method according to claim 6 or 7.
f) A step of setting an inspection area of a tablet based on the reconstructed image, and
g) The step of inspecting the defect of the tablet in the inspection area set by the step f) of the photographed image, and
A tablet printing method that further comprises. The tablet printing method according to claim 8.
In the step f), a plurality of the inspection areas are set based on the reconstructed image, and the inspection areas are set.
In the step g), a tablet printing method for inspecting a tablet defect under different conditions for each inspection area. The tablet printing method according to claim 9.
The tablet printing method, wherein the plurality of inspection areas include an area corresponding to an edge of a tablet.

Images