Nori Making Machine

Abstract

To estimate, from a laver raw sheet before dried, a weight range of dried laver after dried without relying on experiences or hunches of an expert.SOLUTION: A laver producing machine (1) produces dried laver by, after dehydrating a laver raw sheet (19) formed on a laver screen (4), drying the dehydrated laver, in which an estimated weight range of the dried laver is classified using artificial intelligence based on image data (25) of the laver raw sheet (19) before dried. A light emitter (16) and a camera (17) are arranged in facing positions sandwiching the laver screen (4), a specified pattern (18) is arranged between the light emitter (16) and the camera (17), and the image data (25) is acquired using transmitted light of the specified pattern (18), laver screen (4), and laver raw sheet (19). Taking a light-shielding part (28) as a reference value and a laver screen part (22) as a maximum value, values of a specified pattern part (24) and a laver raw sheet part (23) are obtained. The image data (25) is acquired by changing an emitting light color of the light emitter (16).SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a laver manufacturing machine that produces dried laver by dehydrating laver dough that has been made on a laver mat and then drying it. [Background technology]

[0002] Conventionally, in a nori manufacturing machine, a paper making device, a dehydration device, and a drying device are arranged in that order on a conveying path that continuously conveys nori mats. Nori raw material consisting of raw seaweed and water adjusted to a predetermined concentration is supplied to the paper making device, and the paper making device supplies a predetermined amount of the nori raw material to the nori mat, thereby making nori dough of a predetermined shape on the nori mat. The nori dough is then dehydrated in the dehydration device, and then dried in the drying device to produce dried nori.

[0003] The dried seaweed produced by this seaweed manufacturing machine is evaluated for quality based on its weight (density) in addition to taste, color, and gloss, and its commercial value is determined.

[0004] Therefore, conventionally, the weight of dried laver dried in a drying device is adjusted by adjusting the concentration of laver raw material supplied to a paper making device (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Laid-Open No. 2008-109857 Summary of the Invention [Problem to be solved by the invention]

[0006] In the conventional nori seaweed manufacturing machines described above, it takes time from when the nori seaweed dough is made in the paper making device to when it is dried in the drying device to produce dried nori seaweed. Therefore, if the concentration of the nori seaweed raw material to be supplied to the paper making device is adjusted after measuring the weight of the dried nori seaweed dried in the drying device, a large amount of dried nori seaweed will be produced in that time.

[0007] Therefore, in the past, an expert would visually check the condition of the nori dough after it had been made and dehydrated, and adjust the concentration and amount of the nori raw material supplied to the paper-making machine based on the expert's experience and intuition, or based on the weight of the nori dough, which was then removed from the nori-making machine along with the nori mat, dried in a high-speed dryer (such as a microwave oven) installed separately from the nori-making machine, and measured on a scale.

[0008] However, coupled with recent social issues such as a lack of successors and abnormal weather, there is a risk that it will become difficult to continue producing dried seaweed of the desired weight stably by relying solely on the experience and intuition of skilled workers.

[0009] Furthermore, if the seaweed is dried in a high-speed dryer and then weighed on a scale, the accuracy of measuring the seaweed weight can be improved even by an unskilled person, but it takes time to measure one sheet, so it is virtually impossible to measure the weight of all the seaweed. [Means for solving the problem]

[0010] Therefore, in the present invention according to claim 1, in a nori seaweed manufacturing machine that produces dried nori seaweed by dehydrating and then drying nori seaweed dough that has been made into a nori seaweed tray, a light source and a camera are arranged in positions facing each other across the nori seaweed tray, A specific pattern for blocking light from the light emitter only in the area where the specific pattern of a predetermined shape is formed is disposed on the upper part of the light emitter, By taking pictures of the light transmitted through the nori mat and the nori dough before drying, It consists of data of the part that is transmitted only through the nori mat, data of the part that is transmitted through the nori mat and the nori dough, and data of the part that is blocked by a specific pattern. Image data was acquired, and the image data was analyzed using artificial intelligence by extracting features using a CNN (convolutional neural network) using training data created in advance from different raw seaweeds.The weight of the dried seaweed when the seaweed dough was dried was then classified into several estimated weight ranges based on the image data before drying. Furthermore, in the present invention according to claim 2, in the present invention according to claim 1, a light-blocking body is placed between the nori screen and the camera, and the data of the part blocked by the light-blocking body is set as a reference value, and the data of the part that is transmitted only through the nori screen is set as a maximum value, and using these reference value and maximum value, the relative value of the data of the part that is transmitted through the nori screen and the nori dough and the data of the part that is blocked by a specific pattern is calculated and used as image data. [Effects of the Invention]

[0016] The present invention provides the following effects.

[0017] In other words, in the present invention, in a nori manufacturing machine that dehydrates and dries nori dough that has been laid on a nori mat to produce dried nori, artificial intelligence is used to classify the nori dough into estimated weight ranges based on image data of the nori dough before drying, so that the weight range of the dried nori after drying can be estimated from the nori dough before drying without relying on the experience or intuition of an expert.

[0018] In addition, the weight range of all dried seaweed can be estimated on the seaweed production line without removing the seaweed dough from the seaweed manufacturing machine along with the seaweed mat, drying it in a high-speed dryer, and measuring its weight on a scale.

[0019] In particular, if an illuminant and a camera are placed on opposite sides of a seaweed mat, and a specific pattern is placed between the illuminant and the camera, and image data is acquired using light transmitted through the specific pattern, the seaweed mat, and the seaweed dough, or if image data is acquired by changing the light emission color of the illuminant, the weight range of the dried seaweed after drying can be estimated more accurately from the seaweed dough before drying. [Brief explanation of the drawings]

[0020] [Figure 1] (a) is a side view of a seaweed manufacturing machine, and (b) is a top view of a seaweed tray. [Figure 2] FIG. 1 is a side cross-sectional explanatory view showing a seaweed weight estimation device according to a first embodiment. [Figure 3] FIG. [Figure 4] FIG. 4 is an explanatory diagram showing an acquired image. [Figure 5] FIG. 10 is a cross-sectional side view illustrating a seaweed weight estimation device according to a second embodiment. [Figure 6] FIG. [Figure 7] FIG. 4 is an explanatory diagram showing an acquired image. DETAILED DESCRIPTION OF THE INVENTION

[0021] Hereinafter, a specific configuration of the laver manufacturing machine according to the present invention will be described with reference to the drawings.

[0022] As shown in Figure 1, the nori seaweed production machine 1 is configured with a production machine 2 and a dryer 3 arranged in a front-to-back configuration. The production machine 2 and the dryer 3 are provided with intermittent conveying devices 6, 7, and 8 for intermittently conveying a reed frame 5 on which a plurality of nori seaweed screens 4 are arranged in the left-right width direction and detachably stretched.

[0023] The laver screen 4 is transported together with the screen frame 5 by the intermittent conveying device 6 of the manufacturing machine 2 from the upper front end to the upper rear end of the manufacturing machine 2, and then handed over from the intermittent conveying device 6 of the manufacturing machine 2 to the intermittent conveying device 7 above the dryer 3. The laver screen 4 is then transported together with the screen frame 5 by the intermittent conveying device 7 above the dryer 3 from the upper front end to the upper rear end of the dryer 3, then turned back downward at the upper rear end of the dryer 3, transported from the middle rear end to the middle front end of the dryer 3, and handed over from the upper intermittent conveying device 7 above the dryer 3 to the lower intermittent conveying device 8 at the middle front end of the dryer 3. The nori screen 4 is then transported together with the screen frame 5 by the intermittent conveying device 8 below the dryer 3 from the front end of the middle section of the dryer 3 to the rear end of the middle section, then turned back downward at the rear end of the middle section of the dryer 3, transported from the rear end of the lower part of the dryer 3 to the front end of the lower part, and handed over again from the intermittent conveying device 8 below the dryer 3 to the intermittent conveying device 6 of the manufacturing machine 2 at the front end of the lower part of the dryer 3. The nori screen 4 is then transported together with the screen frame 5 by the intermittent conveying device 6 of the manufacturing machine 2 from the rear end of the lower part of the manufacturing machine 2 to the front end of the lower part of the manufacturing machine 2, and then turned back upward at the front end of the lower part of the manufacturing machine 2. As a result, a transport path is formed in the nori manufacturing machine 1 for continuously transporting the nori screens 4 together with the screen frame 5.

[0024] On the conveying path, a paper making device 9 provided in the manufacturing machine 2, a dewatering device 10, a drying device 11 provided in the dryer 3, and a peeling device 12 provided in the manufacturing machine 2 are arranged in this order.

[0025] The nori seaweed manufacturing machine 1 operates intermittent conveying devices 6, 7, and 8, a paper making device 9, a dehydrating device 10, a drying device 11, and a peeling device 12 in an interlocking manner via a driving device 13. A control device 14 is connected to the driving device 13, and the driving of the paper making device 9, dehydrating device 10, drying device 11, and peeling device 12 is controlled by the control device 14 via the driving device 13.

[0026] The Nori producing machine 1 then transports multiple Nori screens 4 along a transport path together with the screen frames 5 using intermittent transport devices 6, 7, and 8, and supplies a predetermined amount of Nori raw material consisting of Nori seaweed and water to the top surface of each Nori screen 4 using the paper making device 9 to make Nori dough of a predetermined shape (for example, square) on the top surface of the Nori screen 4. The Nori dough is then dehydrated in the dehydration device 10, and the Nori dough is then forcibly dried with hot air in the drying device 11 to make dried Nori, after which the dried Nori is peeled off from each Nori screen 4 using the peeling device 12, and the Nori screens 4 together with the screen frames 5 are then transported back to the paper making device 9 again, and these operations are repeated continuously. In this way, the Nori producing machine 1 continuously produces dried Nori from the Nori dough, and moreover, multiple sheets of dried Nori are produced simultaneously.

[0027] In this laver manufacturing machine 1, a laver weight estimation device 15 is provided between the dehydration device 10 (manufacturing machine 2) and the drying device 11 (dryer 3) on the conveyance path.

[0028] The seaweed weight estimation device 15 acquires an image of the seaweed dough before drying in the drying device 11, analyzes the acquired image using artificial intelligence, and classifies the weight of the dried seaweed when the seaweed dough in the acquired image is subsequently dried in the drying device 11 into multiple estimated weight ranges. The seaweed weight estimation device 15 may be controlled by the control device 14 of the seaweed production machine 1, or by a separate, independent control device, or may be an edge device on a network.

[0029] As shown in Figures 2 and 3, in the seaweed weight estimation device 15 of Example 1, an illuminant 16 is placed below the seaweed screen 4 (screen frame 5) transported by the intermittent conveying device 6, while a camera 17 is placed above the seaweed screen 4 (screen frame 5), so that the illuminant 16 and the camera 17 are positioned opposite each other above and below the seaweed screen 4 (screen frame 5).

[0030] In addition, the seaweed weight estimation device 15 has a specific pattern 18 arranged between the light emitting body 16 and the camera 17 to partially block the light emitted from the light emitting body 16.

[0031] The light emitter 16 emits light in a planar manner over an area wider than the nori dough 19 laid on the top surface of the nori mat 4. For example, the light emitter 16 may be configured to use a fluorescent tube or LED as the light source, and a diffuser to emit the light from the light source upward in a planar manner. The light emitter 16 may emit monochromatic light within a predetermined wavelength range, such as red, blue, or green, or may emit a wide variety of wavelengths, such as white light, or may emit a plurality of these lights in a changeable manner. Furthermore, multiple light emitters 16 may be provided, and the light pattern created by the multiple light emitters 16 may be varied.

[0032] The camera 17 photographs an area wider than the nori dough 19 laid on the top surface of the nori screen 4, and acquires image data 25 using light transmitted through the nori screen 4 and the nori dough 19.

[0033] The specific pattern 18 is formed by forming a light-shielding screen 21 of a predetermined shape on a transparent plate 20 wider than the nori dough 19 made on the top surface of the nori screen 4, so that light from the light emitter 16 is blocked only in the area where the specific pattern of the predetermined shape is formed. The light-shielding screen 21 is formed even in an area that extends beyond the nori dough 19 made on the top surface of the nori screen 4.

[0034] As a result, the camera 17 captures the transmitted light that has been emitted from the light emitter 16 and passed through the specific pattern 18, the seaweed mat 4, and the seaweed dough 19, and as shown in Figure 4, image data 25 is obtained that consists of data for the portion that has passed only through the seaweed mat 4 (seaweed mat portion 22), data for the portion that has passed through the seaweed mat 4 and the seaweed dough 19 (seaweed dough portion 23), and data for the portion of the specific pattern 18 that is blocked by the light-shielding screen 21 (specific pattern portion 24).

[0035] A processor 26 (computer) is connected to the light emitter 16 and the camera 17 , and the processor 26 changes the color of light emitted by the light emitter 16 and processes image data 25 acquired by the camera 17 .

[0036] This processor 26 uses artificial intelligence to analyze image data 25 of the dehydrated but undried nori seaweed dough 19, and classifies the nori seaweed dough 19 into estimated weight ranges that are estimated as the weight of the dried nori seaweed dough 19. The estimated weight ranges are preset into multiple categories.

[0037] At this time, the data of the nori mat portion 22, the nori dough portion 23, and the specific pattern portion 24 in the image data 25 are evaluated.

[0038] As shown in FIGS. 5 and 6, in the seaweed weight estimation device 15 according to the second embodiment, a light blocking body 27 is additionally disposed between the seaweed screen 4 and the camera 17.

[0039] 7, data on the nori mat portion 22, the nori dough portion 23, the specific pattern portion 24, and the shading portion 28 (portion shaded by the shading portion 27) are extracted, the data on the shading portion 28 is set as a reference value, and the data on the nori mat portion 22 is set as a maximum value. Using these reference value and maximum value, a relative value is calculated from the data on the nori dough portion 23 and the specific pattern portion 24 to evaluate the image data 25. This makes it possible to correct variations in the image data 25.

[0040] In addition, the light emission color of the light emitter 16 may be changed and the brightness data for each light emission color may be used, or the light emission color of the light emitter 16 may be set to white and the brightness data may be used, or the brightness data of a predetermined color such as red, blue, or green may be used.

[0041] Furthermore, in analysis using artificial intelligence, feature extraction can be performed using CNN (convolutional neural network), and by using data with different colors, origins of seaweed raw materials, harvest times, etc. as training data, it is possible to take into account differences in specific weight. Training data can also be created using transfer learning.

[0042] As described above, the nori manufacturing machine 1 is configured to dehydrate the nori dough 19 that has been laid on the nori mat 4 and then dry it to produce dried nori, and classify the dried nori into an estimated weight range using artificial intelligence based on image data 25 of the nori dough 19 before drying.

[0043] Therefore, in the laver manufacturing machine 1 configured as described above, the weight range of the dried laver after drying can be estimated from the laver dough 19 before drying without relying on the experience or intuition of a skilled person.

[0044] In addition, the nori manufacturing machine 1 is configured such that an illuminant 16 and a camera 17 are positioned opposite each other across the nori mat 4, and a specific pattern 18 is positioned between the illuminant 16 and the camera 17, and image data 25 is acquired using light transmitted through the nori mat 4, the nori dough 19, and the specific pattern 18.

[0045] In addition, a light shielding body 27 can be placed between the nori screen 4 and the camera 17, and image data 25 can be obtained using light transmitted through the nori screen 4, the nori dough 19, the specific pattern 18, and the light shielding body 27, and the light shielding body portion 28 can be used as a reference value and the nori screen portion 22 can be used as a maximum value to obtain values ​​for the nori dough portion 23 and the specific pattern portion 24.

[0046] The specific pattern 18 may be a number, a letter, a symbol, or a shape pattern such as a polygon, a cross, or a circle.

[0047] Therefore, in the laver manufacturing machine 1 configured as described above, the weight range of the dried laver after drying can be estimated more accurately from the laver dough 19 before drying.

[0048] The laver producing machine 1 is configured to acquire image data 25 by changing the luminous color of the light emitter 16.

[0049] Therefore, in the laver manufacturing machine 1 configured as described above, the weight range of the dried laver after drying can be estimated more accurately from the laver dough 19 before drying.

[0050] Furthermore, the nori production machine 1 can perform feedback control to adjust the concentration of the nori raw material supplied by the paper making device 9 based on the results of classification into estimated weight ranges, making it possible to produce dried nori within a predetermined weight range and uniformly produce high-quality dried nori. [Explanation of symbols]

[0051] 1 Nori making machine 2 Making machine 3 Dryer 4 Seaweed cage 5. Screen frame 6,7,8. Intermittent conveying device 9 Paper making equipment 10 Dehydration equipment 11 Drying device 12 Peeling device 13 Drive unit 14 Control unit 15 Seaweed weight estimation device 16 Light source 17 Camera 18 Specific Pattern 19 Nori dough 20 Transparent plate 21 Blackout curtain 22 Seaweed cage part 23 Seaweed dough part 24 Specific pattern part 25 Image data 26 Processing machine 27 Light shielding body 28 Light shielding part

Claims

1. In a nori seaweed manufacturing machine, nori dough is dehydrated and then dried to produce dried nori seaweed. A light-emitting element and a camera are placed on opposite sides of a nori screen, and a specific pattern is placed on top of the light-emitting element to block light from the light-emitting element only in the area where a specific pattern of a predetermined shape is formed. The camera photographs the light transmitted through the nori screen and the nori dough before drying, obtaining image data consisting of data on the area that passed only through the nori screen, data on the area that passed through the nori screen and the nori dough, and data on the area that was blocked by the specific pattern. The image data is analyzed using artificial intelligence by extracting features using a CNN (convolutional neural network) using training data created in advance from different nori raw seaweed, and the weight of the dried nori seaweed after the nori dough is dried is classified into multiple estimated weight ranges based on the image data before drying.

2. A seaweed manufacturing machine as described in claim 1, characterized in that a light-shielding body is placed between the seaweed screen and the camera, and the data of the part blocked by the light-shielding body is used as a reference value, and the data of the part that is transmitted only through the seaweed screen is used as a maximum value, and these reference values ​​and maximum values ​​are used to calculate the relative values ​​of the data of the part that is transmitted through the seaweed screen and seaweed dough and the data of the part that is shielded by a specific pattern, and these are used to obtain image data.

Images