PRINTING SYSTEM, PRINT MEDIUM IDENTIFICATION METHOD, AND MEDIA INFORMATION PROVIDING APPARATUS

Abstract

To output auxiliary information for specifying a type of printing medium and specify the type of printing medium.SOLUTION: A learned model generated by machine learning with an identifier as a teacher is prepared on the basis of spectral information of an unprinted region of a printing medium and the identifier indicating the type of printing medium for the plural types of printing media. On the other hand, the spectral information of the unprinted region of the printing medium to be printed is acquired and the acquired spectral information is applied to the learned model. By using the result, auxiliary information about the type of printing medium is output and specification of the type of printing medium using the output auxiliary information is received. Consequently, printing is performed by using a printing condition suitable to the printing medium when the type of printing medium can be specified.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to techniques for identifying media types for processing the media, such as printing or playing. [Background technology]

[0002] In printing using ink, the type of printing medium (hereinafter referred to as media) such as paper onto which the ink is ejected is identified and the amount of ink applied is adjusted accordingly. In such cases, it is essential to identify the type of media to be printed on. Patent Document 1 generates identification information for identifying the type of media, and based on the identification results using the identification information, displays a list of the corresponding types of media, from which print settings can be selected. If the media is identified as new, it is possible to either register it as new media or select from existing media types. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2006-154913 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, in the technology described in Patent Document 1, in the case of new media, although it can be registered as new media, the corresponding print mode is merely selected from existing modes. Although the registered new media can have its media name changed, the print mode is merely selected and used from existing modes. In other words, even if it is possible to determine that the media is new, it is difficult to link it to an appropriate print mode because there is no information to assist in that determination. Furthermore, when selecting from existing media types, only options are given, and there is no information to determine which existing media is appropriate. For this reason, it is also difficult for the user to select an appropriate existing media type. [Means for solving the problem]

[0005] The present disclosure can be realized in the following forms or application examples.

[0006] (1) A first aspect of the present disclosure is a printing system for printing on a print medium. This printing system includes a storage unit that stores a trained model for a plurality of types of print media, which is generated in advance by machine learning using an identifier indicating the type of the print medium and the spectral information of an unprinted area of ​​the print medium as a teacher, a spectral information acquisition unit that acquires spectral information of the unprinted area of ​​the print medium on which printing is to be performed, an output unit that outputs auxiliary information regarding the type of the print medium using a result of applying the acquired spectral information to the stored trained model, and an identification unit that accepts identification of the type of the print medium using the auxiliary information output by the output unit. and a printing condition setting unit that sets printing conditions suitable for the printing medium when the type of the printing medium is identified based on the auxiliary information output by the output unit, and when the printing medium is not the learned printing medium, the printing condition setting unit allows a user to select whether to print using the printing conditions for the learned printing medium or to create new printing condition settings and perform printing. do.

[0007] (2) A second aspect of the present disclosure is a print medium identification method for identifying a print medium. This print medium identification method prepares a trained model for a plurality of types of print media, which is generated by machine learning using the identifier as a teacher, based on spectral information of an unprinted area of ​​the print medium and an identifier indicating the type of the print medium, acquires spectral information of the unprinted area of ​​the print medium to be printed on, uses a recognition result obtained by applying the acquired spectral information to the trained model to output auxiliary information regarding the type of the print medium, and accepts identification of the type of the print medium using the output auxiliary information. When the type of the print medium is specified based on the outputted auxiliary information, printing conditions suitable for the print medium are set, and if the print medium is not the learned print medium, the user is prompted to select whether to use the printing conditions for the learned print medium or to create and use new printing condition settings when printing. .

[0008] (3) A third aspect of the present disclosure is a media information providing device. This media information providing device includes: a storage unit that stores a trained model generated by machine learning using a plurality of types of media, for which spectral information of the background color of the media and an identifier indicating the type of the media are given, and the identifier is used as a teacher; a spectral information acquiring unit that acquires spectral information of the background color of the media to be processed; and an output unit that outputs auxiliary information regarding the type of the media by utilizing a result of applying the acquired spectral information to the trained model. and a printing condition setting unit which, when the type of the medium is identified based on the auxiliary information output by the output unit, sets printing conditions suitable for the medium, and when the medium is not the learned medium, the printing condition setting unit allows a user to select whether to use the printing conditions for the learned medium or to create and use new printing condition settings when printing. do. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a schematic configuration diagram showing a printing system according to an embodiment. [Diagram 2] 13 is a flowchart showing a trained model generation process. [Diagram 3] FIG. 1 is a schematic diagram of a supervised machine learning device. [Figure 4] 4 is a flowchart showing a print processing routine. [Diagram 5] 10 is a flowchart showing a known media print profile setting process. [Figure 6] FIG. 13 is an explanatory diagram illustrating a dialog box for selecting a profile for media in the cloud. [Figure 7] FIG. 11 is an explanatory diagram illustrating a dialog box for creating a profile. [Figure 8] 10 is a flowchart showing an unknown media print profile setting process. [Figure 9] FIG. 11 is an explanatory diagram showing an example of display of media types and similarities. [Figure 10] FIG. 13 is an explanatory diagram showing an example of a dialogue box displayed when a medium is determined to be a new medium. [Figure 11] FIG. 13 is an explanatory diagram showing another example of the dialogue box displayed when a medium is determined to be a new medium. [Figure 12] FIG. 11 is a schematic configuration diagram of a media information providing device according to a second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] A. First embodiment: (1) Printer 20 configuration: FIG. 1 is a block diagram showing the configuration of a printing system 10 according to an embodiment. The printing system 10 includes a printer 20 and a machine learning device 50. The printer 20 and the machine learning device 50 are connected via a network NT. The network NT may be wired or wireless. The printer 20 and the machine learning device 50 may be connected via a public network such as the Internet, or may be directly connected. In this embodiment, the machine learning device 50 is realized as a device on a so-called cloud.

[0011] The printer 20 includes a processor 21, a communication unit 23, a display 24, a non-volatile memory 30, a head transport device 25, a paper transport device 26, a print head 27, and the like. The processor 21 functions as a print control unit 22 that controls the entire printer 20 to realize printing. The print control unit 22 is capable of exchanging data with the machine learning device 50 via the communication unit 23. The print control unit 22 can also access the non-volatile memory 30 and refer to a trained model 31 and a print profile 33, which will be described later. The print control unit 22 functions as a print medium recognition unit that uses the trained model 31 to recognize the type of print medium.

[0012] In this embodiment, the printer 20 includes a processor 21 equipped with a CPU, RAM, etc., and a non-volatile memory 30, and functions as a print control unit 22 are realized by the processor 21 executing various programs, such as a print control program, recorded in the non-volatile memory 30. The print control unit 22 displays information required for printing on a display 24. A touch panel is provided on the display 24, and the user can refer to the displayed information and, in some cases, give instructions for operations or select parameters. These processes are also realized as functions of the print control unit 22 by executing programs prepared in advance.

[0013] The print control unit 22 uses the head transport device 25 to reciprocate the carriage 28 carrying the print head 27 in the width direction (main scanning direction) of the print medium P. Furthermore, the print control unit 22 can transport the print medium P in a direction (sub-scanning direction) intersecting the main scanning direction using the paper transport device 26. The carriage 28 can be equipped with ink cartridges 41 each containing ink of four colors, cyan C, magenta M, yellow Y, and black K. The ink in the ink cartridges 41 is supplied to nozzles provided in the print head 27 and ejected onto the print medium P in response to an instruction from the print control unit 22. Therefore, the print control unit 22 can form an image on the print medium P by ejecting ink from each nozzle of the print head 27 in accordance with the transportation of the print medium P in the sub-scanning direction and the reciprocating movement of the print head 27 in the main scanning direction.

[0014] In this embodiment, the printer 20 is a so-called inkjet serial printer, but it may be a printer other than the inkjet type, for example a so-called page printer such as a laser printer. Alternatively, it may be a thermal dye-sublimation printer, and any printer can be used as long as it performs printing by setting different printing conditions according to the type of print medium P, for example, in the case of an inkjet type, the upper limit of the amount of ink applied (ink duty) for each print medium P, the drying temperature if a heater for drying the ink is provided, and further the transport speed of the print medium P.

[0015] The carriage 28 is provided with a spectroscopic reflectometer 43 as a spectral information acquisition unit. The spectroscopic reflectometer 43 is provided with an LED that irradiates light in the wavelength range of visible light and a sensor that detects the intensity of the light from the LED reflected by the printing medium P in a predetermined wavelength range, for example, every 10 nm. The spectroscopic reflectometer 43 is provided at the leading end of the carriage 28 in the sub-scanning direction, so that it can detect the spectral reflectance of the unprinted portion at the leading end of the transported printing medium P prior to printing. In addition to the spectral reflectance, the spectral information may also be the spectral transmittance, which is the rate at which light passes through the printing medium P. In this case, a light-emitting LED and a light-receiving sensor may be provided at a position that sandwiches the printing medium P.

[0016] The spectral reflectance data acquired from the spectroscopic reflectometer 43 is used to identify the type of printing medium P. In this embodiment, the print control unit 22 recognizes the type of printing medium P by providing the spectral reflectance data detected using the spectroscopic reflectometer 43 to the trained model 31 stored in the non-volatile memory 30. A method for recognizing the type of printing medium P using the spectral reflectance will be described in detail later. When the type of printing medium P is recognized, the print control unit 22 reads out the printing profile 33 corresponding to the type of printing medium P to be printed from the non-volatile memory 30. The printing profile 33 usually includes multiple types of parameters, and is therefore prepared as a set of multiple printing parameters corresponding to the type of printing medium P.

[0017] In this embodiment, the type of print medium P is identified by an identifier number indicating the type of print medium P and a parameter indicating that the print medium P is that type of print medium by providing spectral reflectance data to the trained model 31. The identifier number indicating the type of print medium P and the parameter indicating the possibility that the print medium P is that type of print medium are referred to as auxiliary information regarding the type of print medium.

[0018] The trained model 31 is created by supervised machine learning, and is created in advance and installed in the non-volatile memory 30 of the printer 20. This trained model 31 is generated by deep learning in the machine learning device 50. The trained model 31 generated by the machine learning device 50 is downloaded to the printer 20 via the communication unit 23 prior to printing.

[0019] (2) Configuration of machine learning device 50: As shown in the figure, the machine learning device 50 placed on the cloud includes a learning control unit 51, a communication unit 53, a machine learning unit 60, a non-volatile memory 70, and a bus 55 connecting these. The communication unit 53 communicates with the communication unit 23 of the printer 20 to exchange data with the printer 20.

[0020] The machine learning unit 60 includes a dedicated processor and memory for performing supervised machine learning. The memory stores a training model 61, teacher data 63, and a dedicated learning program, and by executing the learning program, the training model is trained using the teacher data 63 to generate a trained model 65. The trained model 65 is a trained model generated using the training model 61 with the teacher data obtained up to a certain point in time, and is treated as the trained model 31 in the printer 20 when downloaded to each printer 20 via the communication unit 53. At the time of downloading, the trained model 65 of the machine learning device 50 and the trained model 31 in the printer 20 are the same.

[0021] In this embodiment, the machine learning device 50 performs machine learning using a training model 61 with the spectral reflectance data of various print media P and an identifier indicating the type of print medium P corresponding to the spectral reflectance data as teacher data 63. This utilizes the fact that the characteristics of the spectral reflectance differ depending on the type of print medium P. In the case of white paper that does not use a fluorescent whitening agent, there tends to be little change in reflectance in each wavelength range. On the other hand, in the case of white paper that uses a fluorescent whitening agent, there is a tendency for the reflectance to increase in the short wavelength range. In this way, the spectral reflectance of the print medium P includes information such as color information of the surface of the print medium P and the presence or absence of a fluorescent whitening agent.

[0022] 2 is a flowchart showing the process in which the machine learning unit 60 performs machine learning to generate a trained model. Before starting this process, training data is collected for multiple types of print media P, in which the spectral reflectance data is associated with an identifier that identifies the type of print medium P. Because the process of generating the trained model is performed based on this training data, the machine learning device 50 and the printer 20 do not need to be connected when the process of generating the trained model is performed.

[0023] When the process starts, a process of acquiring teacher data 63 is performed first (step S100). The teacher data 63 is collected in advance prior to the process. In this embodiment, the teacher data, which is the spectral reflectance data, is collected for 20 types of media from media A to media T and stored in memory. After the teacher data 63 is acquired, machine learning processing (step S110) using the training model 61 is performed until generalization is completed (step S120). The teacher data 63 is given to the training model 61 and machine learning is performed, thereby generating a trained model 65, which is finally stored as the trained model 65. The stored trained model 65 is transferred to the printer 20 when the printer 20 is connected to the machine learning device 50 and stored in the non-volatile memory 30.

[0024] A training model for machine learning can be defined in various ways. FIG. 3 is a diagram showing an example of a model used in this embodiment. In the figure, each layer of the total n layers by CNN is shown as a rectangular parallelepiped L1 to Ln, and the nodes of a normal neural network are shown as white circles. In this embodiment, a CNN (Convolutional Neural Network) is used for the data of the spectral reflectance, but other models such as various neural networks such as a capsule network type and a vector neural network type may be used. As an example, a configuration using a vector neural network shown in Japanese Patent Application No. 2020-94205, which is another application by the applicant of the present application, can be mentioned. In this vector neural network type model, the output of the intermediate layer can be used. Note that vector neural networks that can use the output of the intermediate layer are publicly known.

[0025] The first layer L1 is provided with a plurality of nodes that input reflectance for each fixed wavelength. In this embodiment, the spectral reflectance for each fixed wavelength indicated by the spectral reflectance data is input as input data to each node of the first layer L1, which is the input layer, and final output data related to the spectral reflectance data is output from the final output layer Ln. The reflectance is a value between 0 and 1. Instead of the spectral reflectance data, it is also possible to use transmittance for each fixed wavelength. It is also possible to use both. When using both the spectral reflectance and the spectral transmittance, two training models as shown in FIG. 3 are provided, the results of the spectral reflectance and the spectral transmittance are output from the final layer of each model, and a final output layer Ln that integrates and judges both results may be constructed to output the final result.

[0026] The output of each node in the first layer L1 is connected to a node in the next second layer with a predetermined weighting. This is the same for the second layer L2 and onwards, up to the n-1th layer. By repeating the process of correcting the weighting between each node in each layer using training data, learning is completed and a trained model 65 is generated. This point will be explained in detail below.

[0027] In learning, the machine learning unit 60 inputs one of the teacher data into the training model and calculates the output value of the final output layer Ln. At this time, it is determined in advance which node of the final output layer Ln corresponds to which type of media. For example, if the print medium P is medium A, learning is performed so that node A outputs a value close to value 1. One teacher data is given to the training model 61, and the machine learning unit 60 evaluates the value of each node A to T of the output final output layer Ln by a loss function. In other words, if the teacher data is for, for example, medium A, a state in which node A has a value of 1 and other nodes B to T have a value of 0 is set as the target state, and the difference from the actual output (learning error) is specified by the loss function. Then, the machine learning unit 60 corrects the weighting between each node. After the correction, the learning error is calculated again by the loss function, and the difference (differential value) from the previously calculated learning error is calculated, and the weighting is further corrected based on whether the learning error has been reduced by the weighting.

[0028] The loss function for determining the learning error can be various functions, for example, cross entropy error. The process of calculating the loss function as described above is performed on all or part of the data included in the teacher data 63, and the loss function for one learning session is expressed by the average or sum of the data. Once the learning error for one learning session is obtained, the machine learning unit 60 modifies the weighting using a predetermined optimization algorithm, for example, stochastic gradient descent.

[0029] When the weighting correction is performed a preset number of times for all the teacher data 63 in this manner, the machine learning unit 60 determines whether generalization of the training model 61 is complete (step S120). That is, the machine learning unit 60 obtains the degree of agreement between the input and output of the training model 61 using the teacher data 63, and determines that generalization is complete when the degree of agreement is equal to or greater than a predetermined threshold.

[0030] If it is not determined that generalization is complete, the machine learning process (step S110) is repeated. That is, a process of further modifying the weights is performed, and the calculation and evaluation of the learning error using the error function is repeated. In addition to the supervised machine learning of the type of print medium P, the machine learning unit 60 may also perform other processes, such as verifying the validity of hyperparameters.

[0031] Each node in the final output layer Ln corresponds to the type of print medium P to be discriminated (media A, media B, media T). Using the training data, the sum of the output values ​​of each node in the output output layer Ln when one training data is given is normalized to be 1. It is possible to estimate that the type of print medium P corresponding to the node with the largest output value in the final output layer Ln is the type of print medium P from which each training data input to the input layer L1 was obtained. Ideally, when each training data is given, the output of one node in the final output layer Ln is value 1 and the others are value 0, and it is desirable to be able to discriminate the print medium P, but the output of each node in a multi-layered neural network with multiple types of output is usually in the range greater than value 0 and less than value 1. For this reason, the maximum value of the output values ​​of each node in the final output layer Ln is treated as a parameter indicating the possibility that the print medium P is the media corresponding to the node.

[0032] In this way, in the machine learning device 50, the training model 61 is given the teacher data 63 to learn, and a trained model 65 is generated. For the type of print medium P for which supervised machine learning is performed, a print profile 73 for printing on that type of print medium P is prepared and stored in the non-volatile memory 70. Since the print profile 73 differs depending on the printer 20, it is assumed that it may be difficult to store the print profile 73 for all types of print media P on the printer 20 side. For this reason, for some types of print media P, the corresponding print profile 73 is stored in the non-volatile memory 70 of the machine learning device 50 on the cloud side. Of course, a dedicated site for storing the print profile 73 may be provided on the cloud, and the print profile 73 may be prepared there.

[0033] The trained model 65 generated by the machine learning device 50 is downloaded to each printer 20 as necessary, and becomes usable in each printer 20. At this time, the printer 20 also downloads the trained data of the spectral reflectance for each type of media together with the trained model 65, and stores this in the non-volatile memory 30. When a new medium appears, the machine learning device 50 performs machine learning using the medium as supervised data with an identifier attached, so that the trained model 65 in the machine learning unit 60 is always updated to the latest one. On the other hand, the trained model stored in the non-volatile memory 30 of each printer 20 becomes the latest one at the time each printer 20 downloads it, and the two do not necessarily match until then. For this reason, the trained model used by each printer 20 is indicated by the reference symbol 31, and is distinguished from the trained model 65 generated in the machine learning unit 60.

[0034] (3)Printing process: FIG. 4 shows a print processing routine executed by the printer 20. Here, when the printer 20 receives a print instruction from a computer, mobile phone, or the like (not shown), the printer 20 performs both the media discrimination process (steps S200 to S400) and the print process (step S250). However, the media discrimination process and the print process may be performed separately. If the print medium P is roll paper or the like, the media discrimination process may be performed when the print medium P is loaded into the printer 20. In the case of cut sheets, the media discrimination process may be performed on the first sheet, and thereafter, the media discrimination process may be skipped until the print medium P is replaced. Before the process routine shown in FIG. 4 is started, the non-volatile memory 30 stores the learned model 31 and the print profile 33 corresponding to multiple types of print media P. When the process of FIG. 4 starts, the processor 21 of the printer 20 performs a process of transporting the print medium P prepared in a tray (not shown) or the like until the leading edge of the print medium P is positioned under the spectroscopic reflectometer 43 (step S200).

[0035] Next, a process of acquiring the spectral reflectance of the printing medium P is performed using the spectroscopic reflectometer 43 (step S210). The wavelength of the irradiated light is divided into a predetermined wavelength range, here every 10 nm, and the spectral reflectance is acquired as a set of data on the reflectance in the wavelength range. In FIG. 4, the solid line J1 shows an example of the spectral reflectance of normal printing paper. In contrast, the printing medium P to which a fluorescent whitening agent is applied has a higher reflectance on the short wavelength side than normal printing paper, as shown by the solid line J2. The spectral reflectance may be measured by the printer 20, or may be measured using a dedicated measuring device, and the processor 21 may acquire the reflectance via the communication unit 23 using communication or the like.

[0036] Next, processor 21 performs a media determination process to determine the type of print medium P using the spectral reflectance data acquired in step S210 and trained model 31 (step S220). Processor 21 inputs the acquired spectral reflectance data to trained model 31 and obtains an output value. Processor 21 acquires an identifier and parameters representing the type of print medium P based on the output values ​​of the intermediate layer and final layer of trained model 31 (step S230).

[0037] Next, based on the acquired parameters, it is determined whether the print medium P to be used for printing is a known medium (step S240). Of course, the parameters obtained from the output values ​​of the intermediate layer and the final output layer Ln and the type (identifier) ​​of the medium may be displayed, and the user may be asked to confirm whether it is OK to determine that the medium is a known medium. In this case, it is up to the user to determine whether the medium is a known medium.

[0038] If it is determined that the medium is a known medium, the known medium print condition setting process (step S300) is executed, and if it is determined that the medium is not a known medium, the unknown medium print condition setting process (step S400) is executed. Either print condition setting process sets various parameters required for printing, and printing is executed using these print parameters (step S250).

[0039] The known media printing condition setting process (step S300) will be described with reference to Figs. 5 to 7. When the print medium P set in the printer 20 is determined to be a known medium by the media determination process (step S220) using the trained model 31, the media name corresponding to the acquired identifier is displayed (step S305). The media name may be displayed on the display 24 provided on the printer 20, or on the display unit of the computer or mobile phone that issued the printing command. Hereinafter, "display" refers to any of these several display modes.

[0040] Next, it is determined whether a printing profile for the known medium with the media name displayed is local, i.e., in the printer 20 (step S310). The printing profile 33 is a set of multiple printing parameters corresponding to the type of printing medium P. (A) Basic LUT (table for converting digital values ​​into ink amounts) determined by print settings (resolution, print medium) (B) ICC profile The main types of print media P, such as plain paper and matte photo paper, are stored in the non-volatile memory 30 in advance. In this case, the processor 21 sets the print profile 33 stored in the non-volatile memory 30 (step S320), and exits to "NEXT" to end this processing routine. After this, printing is performed (FIG. 4, step S250). In this embodiment, the printer 20 is configured to have the print profile 33, considering the case where a memory card is inserted in the printer 20 and printing is performed, or a print instruction is received from a terminal such as a mobile phone that does not have a printer driver, but when printing from a computer or the like that has a printer driver, the print profile 33 is set on the computer side together with the printer driver installed in the computer. Therefore, in step S310, it is sufficient to determine whether a print profile for known media that displays the media name is present on the printer driver side.

[0041] On the other hand, if it is determined in step S310 that the print medium P loaded in the printer 20 is a known medium but that its print profile 33 is not locally available, then a selection dialog box is displayed (step S330). An example of the selection dialog box 90 is shown in FIG. 6. In this example, media A through N are prepared in the non-volatile memory 30 of the printer 20, but media O through T are not prepared. For media that are not prepared, a print profile 33 may be prepared in the cloud, so if the print medium P determined to be a known medium corresponds to media O through T, an option to select one of the profiles in the cloud and an option not to use a profile in the cloud are displayed.

[0042] Then, a process of accepting input from the user is performed using the selection dialog box 90 (step S340). As long as the selection dialog box 90 is displayed, other processes cannot be accepted. Therefore, when the user selects one of the options by touch panel or mouse click, presses the "OK" button, and closes the selection dialog box 90, the processor 21 judges whether the user's selection is to use a profile in the cloud (step S350). If the user's selection is to use a printing profile in the cloud, the machine learning device 50, which is a cloud on the network NT, is accessed via the communication unit 23, and the corresponding printing profile 73 is obtained from the non-volatile memory 70 (step S360). After obtaining the printing profile, the printing profile 73 is set (step S320) in the same way as when the printing profile 73 is locally stored, and the process exits to "NEXT" to end this processing routine. Then, printing is performed as described above.

[0043] When the selection dialog box 90 shown in FIG. 6 is used to select "Do not use a print profile in the cloud" (step S350, "NO"), a process for creating a profile for that print medium P is performed (step S370). A new print profile 73 is created using, for example, the selection dialog box 91 shown in FIG. 7. In the figure, the ink duty, the set temperature of the drying heater, the transport speed of the print medium P, and the interlace method are exemplified as configurable items. Ink duty indicates the maximum amount of ink that can be applied to the print medium P in an inkjet printer. The temperature of the drying heater sets the temperature at which the heater is used when a heater is provided to dry the ink ejected from the print head 27. In the printer 20 of this embodiment, a heater is not provided, so this item cannot be set. Items that cannot be set may be not displayed or displayed in a different mode from the other options, such as gray. The transport speed is the speed at which the print medium P is transported using the paper transport device 26. The item for the interlace method sets the order and combination of dot formation when forming an image.

[0044] Once the above settings have been made and the OK button is pressed, the processor 21 uses the basic LUT included in the printing profile 33, utilizing the ink duty set in Figure 7, and then performs printing colorimetry to create an ICC profile for the new printing medium.

[0045] Processor 21 registers this, together with the media name, as one of the printing profiles in non-volatile memory 70, which stores a database of printing profiles 73 (step S380). At this time, information on the conditions for creating the ICC profile is also saved. The reason for registering the created printing profile in non-volatile memory 70 is so that it can be reused when a similar printing medium P is set in printer 20. Subsequent settings of the printing profile (step S320) and printing execution in the printing process routine (FIG. 4, step S250) are the same as those described above.

[0046] The above describes the process when it is determined that the media is a known medium based on the parameters acquired using the trained model 31. On the other hand, when it is determined that the media is not a known medium based on the acquired parameters but an unknown medium, unknown media printing condition setting process is executed. An overview of this process is shown in the flowchart of FIG. 8. When it is determined in step S240 of FIG. 4 that the print medium P set in the printer 20 is an unknown medium, the processor 21 first performs a process to prompt the user to input a new media name (FIG. 8, step S410). The name of the new media can be input by using a touch panel to select a button with letters displayed on the display 24, or from the computer or mobile phone that has instructed to print the image.

[0047] When the user inputs the name of the new media, the system then performs a process of displaying the similarity with the existing media using the learned spectral reflectance data of each medium downloaded from the machine learning device 50 and the currently acquired spectral reflectance data of the print medium P (step S420). An example of the display of the similarity is shown in Fig. 9. Here, the similarity for the already learned media A to M is calculated by the following formula (1) and displayed as a listable table 95. R = 1.0 - ΔE / NR … (1) Here, ΔE is the color difference calculated from the spectral reflectance of both, and NR is a coefficient for normalizing the similarity R from a value of -1.0 to a value of +1.0. Note that when R is a value of -1 or less, R is set to -1.0.

[0048] The closer the similarity R between print media P is to a value of 1.0, the closer the optical characteristics of the print media P are to each other, and the closer the similarity R is to a value of -1.0, the more the optical characteristics are dissimilar. After displaying this similarity R, processor 21 performs processing to display a selection dialog box 97 that allows the user to specify which media to select as similar media (step S430). An example of this selection dialog box 97 is shown in FIG. 10. Using this selection dialog box 97, the user selects whether to select a printing profile 33 from an existing medium with reference to the similarities shown in FIG. 9 for an unknown medium, i.e., a new medium, or to create a printing profile 33 for the new medium by making completely new print settings.

[0049] In the selection dialog box 97, identifiers indicating the print media P with high similarity and the similarity calculated by the above formula (1) are displayed in descending order of the similarity value. In the selection dialog box 97, five types of identifiers (media types in the figure) are displayed, but a scroll bar is provided, so that the remaining media types and their similarities can be confirmed by scrolling through the column showing the media types in descending order of similarity. Even if it is determined that the print medium P set in the printer 20 is not an existing medium, the user may select a media type with a certain degree of similarity based on the similarity and the actual appearance of the print medium P. In this case, the user looks at the display in the selection dialog box 97 and usually selects the media type with the highest similarity and presses the "OK" button. Of course, even if the similarity is low, the user may also select a desired media type based on his or her own experience, even if the similarity is not the highest.

[0050] In FIG. 10, the existing media types are displayed in descending order of similarity, but the order may be reversed, or, for example, the order of the media type names may be used, or they may be displayed randomly. Furthermore, in addition to similarity, the media types may be ranked in advance based on various conditions and displayed together with the rank. An example of this is shown in FIG. 11. In the illustrated selection dialogue box 98, the media types are displayed according to categories such as ranks S, A, and B. The ranks may correspond to similarity, or the print medium P for which the trained model 31 shows the highest parameters may be ranked S, and media with paper quality and ink absorption similar to that of the print medium P may be ranked higher, such as rank A.

[0051] When an existing medium is selected through the selection dialog box 97 or 98 (step S440: "YES"), the printing profile 33 corresponding to the selected media type is selected from the existing printing profiles 33 stored in the non-volatile memory 30 (step S450), and media learning is performed based on this (step S455). Media learning is a process of re-learning the spectral reflectance data of this printing medium P acquired using the spectroscopic reflectometer 43 as being for the selected media type. The actual processing is performed by the machine learning device 50. That is, the processor 21 outputs the type of the existing media received through the selection dialog box 97 or 98 and the spectral reflectance data acquired using the spectroscopic reflectometer 43 to the machine learning device 50 through the communication unit 23, and causes the machine learning unit 60 of the machine learning device 50 to perform re-learning.

[0052] When re-learning is performed by the machine learning unit 60, the database of the printing profile 73 in the non-volatile memory 70 is also updated in response to the results of the re-learning. Then, upon receiving a notification that the re-learning is complete, the processor 21 of the printer 20 downloads the trained model 65 to the non-volatile memory 30 as the trained model 31 to be used by the printer 20, as necessary. Also, the combination of the type of printing medium P and the printing profile to be used for that printing medium P is downloaded from the database of the printing profile 73 in the non-volatile memory 70 to the non-volatile memory 30 (step S460). Therefore, the trained model 31 and the printing profile 33 obtained by the re-learning can be used in the printer 20 thereafter.

[0053] With the selection dialog boxes 97 and 98 displayed, the user can also select to make new print settings without using the print profile for an existing media type. In this case, the determination in step S440 is "NO", and the processing from step S470 onwards is executed. In this case, instead of selecting and using an existing profile (step S450), a new profile is created (step S470). The new print profile (ICC profile) is created by reusing the settings for the print medium P with the highest similarity. Note that, as explained with reference to FIG. 7, the user sets the various parameters that make up the print profile himself.

[0054] After creating a new printing profile, since it is determined that the medium is unknown, an instruction to re-learn the medium (step S475) and download from the machine learning device 50 (step S480) are given, just as in the case of selecting an existing profile (step S450). These processes are the same as the processes in steps S455 and S460, respectively.

[0055] After the above-described processing (steps S440 to S480), a print profile corresponding to the new medium is set, and the process proceeds to "NEXT" to execute printing in the print processing routine (FIG. 4, step S250).

[0056] According to the printing system 10 of the embodiment described above, when printing on the print medium P, the spectral reflectance of the unprinted areas of the print medium P is measured by the spectroscopic reflectometer 43, and this is applied to the trained model 31 to identify the type of the print medium P and then print on the print medium P. If the print medium P cannot be uniquely identified, auxiliary information regarding the type of print medium P, such as the similarity R, is output and displayed, allowing the user to easily identify the type of print medium P. Following this identification, a print profile corresponding to the type of print medium is applied, and printing on the print medium P can be performed under appropriate printing conditions.

[0057] In the embodiment, the similarity R for each type of print medium, which is auxiliary information, is displayed when the print medium P to be printed is an unknown medium. This allows the user to know which known medium the unknown medium is similar to, and allows the machine learning unit 60 to re-learn the print medium using the existing media as a teacher.

[0058] Furthermore, in this embodiment, the type of the printing medium P is identified or auxiliary information is output only based on the spectral reflectance of the unprinted area of ​​the printing medium P, which has the advantage of simplifying the device configuration. Moreover, since the spectroscopic reflectometer 43 is provided in the print head 27, the spectral reflectance of the printing medium P can be obtained by a normal printing method in which the printing medium P to be printed is set in the printer 20. Therefore, it is not necessary to take the troublesome work of bringing the printing medium P to a measuring device and measuring it every time. Furthermore, in this embodiment, since the printer 20 is connected to the machine learning device 50 via the network NT, if the printing medium P is a medium whose type cannot be identified, it is possible to perform supervised machine learning again and update the learned model, printing profile, etc. in real time.

[0059] In this embodiment, the supervised machine learning is performed by a machine learning device 50 separate from the printer 20, so there is no need to prepare a device configuration for performing machine learning in the printer 20. Moreover, in this embodiment, the trained model 65 trained by the machine learning device 50 is downloaded to the non-volatile memory 30 of the printer 20 and can be used by the printer 20 alone. Therefore, there is no need to exchange data with the machine learning device 50 each time the type of printing medium P is identified using the spectral reflectance data. Although FIG. 1 shows the printing system 10 in which the printer 20 and the machine learning device 50 are connected, if re-learning of the trained model is not performed, operation with only the printer 20 is also possible.

[0060] B. Second embodiment: The second embodiment is in the form of a media information providing device 100 shown in FIG. 12. This media information providing device 100 has a configuration including a media processing unit 180, except for the configuration for printing in the printer 20 of the first embodiment. This media information providing device 100 provides auxiliary information that assists in identifying the type of media M. This media information providing device 100 includes a control unit 110 as a media recognition unit that is given spectral information of the background color of the media and an identifier indicating the type of the media for multiple types of media M and recognizes the type of media with predetermined parameters using a trained model 131 generated by machine learning using the identifier as a teacher, a transport unit 126 that transports the media M, a spectral information acquisition unit 143 that acquires spectral information of the background color of the transported media, an output unit 124 that inputs the acquired spectral information to the control unit 130 as the media recognition unit and uses the recognition result by the trained model 131 to output auxiliary information regarding the type of media M, and a processing unit 180 that processes the media M whose type has been identified using the auxiliary information under appropriate processing conditions according to the type. The trained model 131 is stored in memory 130 together with a processing profile 133 that records the processing details as necessary for each type of media M. Note that the media information providing device 100 does not necessarily require the transport unit 126 or the media processing unit 180. Instead of the transport unit 126, the user may bring the media to the spectral information acquisition unit 143 or the media processing unit 180. There may also be cases where the spectral information acquisition unit 143 or the processing unit 180 are not built-in, and the processing is carried out in another factory or device.

[0061] According to the media information providing device 100, auxiliary information for determining the type of media based on the spectral information of the background color of the media is output, so that the type of the media M to be processed can be easily determined. The media information providing device 100 has the advantage that, for example, when recycling paper used for printing, the type of media can be easily determined and suitable conditions can be processed. This is because, when recycling media such as printing paper, it is often necessary to change the processing method or processing procedure depending on the type of media. Recycling media such as print media may require processing steps such as cutting the media or melting it with chemicals, and in such cases, it is useful to be able to easily identify the type of media. In such cases, the media information providing device 100 outputs auxiliary information regarding the type of media based on the spectral information of the background color of the media, so that handling such as processing of the media by the media processing unit 180 can be made easier.

[0062] C. Other embodiments: The printing system of the present disclosure can also be implemented in the following manner. (1) A first aspect is a printing system that prints on a print medium, and includes a storage unit that stores a trained model that is generated in advance for a plurality of types of print media by machine learning using an identifier indicating the type of the print medium and spectral information of an unprinted area of ​​the print medium as a teacher, a spectral information acquisition unit that acquires spectral information of the unprinted area of ​​the print medium on which printing is to be performed, an output unit that outputs auxiliary information regarding the type of the print medium using a result of applying the acquired spectral information to the stored trained model, and an identification unit that accepts identification of the type of the print medium using the auxiliary information output by the output unit.

[0063] According to this printing system, auxiliary information regarding the type of printing medium is output by applying spectral information of unprinted areas of the printing medium to the learned model, allowing the user to easily identify the type of printing medium.

[0064] (2) In such a printing system, the spectral information may be at least one of the spectral reflectance and the spectral transmittance, and the spectral information acquisition unit may include a measurement unit that measures at least one of the spectral reflectance and the spectral transmittance of the unprinted area of ​​the printing medium. In this way, it is only necessary to acquire the spectral reflectance of the printed horse body, and the configuration of the spectral information acquisition unit can be simplified. Spectral transmittance may be used as the spectral information. In addition, both the spectral reflectance and the spectral transmittance may be used. In general, in the case of spectral reflectance, a light projecting unit that projects light and a light receiving unit that receives reflected light can be arranged together on one side of the printing medium, so that the configuration can be simplified. Note that a device that actually performs measurements, such as a spectroscopic reflectometer that detects the spectral reflectance, may be incorporated into the printing system, such as the printing unit, or may be provided separately from the printing system, and the spectral information acquisition unit may acquire the spectral reflectance via communication or the like.

[0065] (3) In such a configuration, the storage unit may include a re-learning unit that re-learns a trained model for a print medium that was previously determined to be an unknown print medium by the trained model using spectral information of the unknown print medium and an identifier assigned to the unknown print medium, and an update unit that updates the stored trained model using a learning result by the re-learning unit. In this way, the trained model based on machine learning can be adapted to new types of print media.

[0066] (4) In such a configuration, a printing condition setting unit may be provided that sets printing conditions suitable for the unknown printing medium for the printing medium determined to be the unknown printing medium. In this way, printing conditions suitable for a new type of printing medium can be set.

[0067] (5) The apparatus may further include a printing condition setting unit that sets the printing conditions suitable for the printing medium when the type of the printing medium is identified based on the auxiliary information output by the output unit. When the type of the printing medium is identified using the result of application to a trained model, the printing condition setting unit can set printing conditions suitable for the identified type of printing medium.

[0068] (6) In such a configuration, the output unit may include a display unit that displays the auxiliary information, and the printing condition setting unit may include an identification unit that allows a user to identify the print medium using the auxiliary information displayed on the display unit. Since the auxiliary information is displayed on the display unit, the user can easily identify the type of print medium. The display unit may be provided on the printer side that performs printing, or on the side of a terminal such as a computer or mobile phone that issues a print instruction.

[0069] (7) In such a configuration, if the printing medium is not the learned printing medium, the printing condition setting unit may allow the user to select whether to print using the printing conditions for the learned printing medium or to create new printing condition settings and print. In this way, the user can know the degree to which the new type of printing medium is similar to the learned printing medium, making it easy to identify the type of printing medium. Note that such identification of the new type of printing medium includes not only uniquely determining the type of printing medium, but also identifying which existing printing medium the new type of printing medium is similar to. In such a case, machine learning may be performed using an identifier indicating the type of similar printing medium as a teacher, and the learned model may be re-trained.

[0070] (8) In this configuration, if the print medium is not the learned print medium, the printing condition setting unit may allow the user to select whether to print using the printing conditions for the learned print medium or to create new printing condition settings and print. In this way, it is possible to appropriately determine whether to set new printing conditions.

[0071] (9) In this configuration, when the creation of the new printing condition setting is selected, the output unit may display the similarity calculated from the spectral information on the display unit as information indicating the degree to which the print medium determined to be unlearned is similar to the learned print medium as one of the auxiliary information. In this way, the degree of similarity of the types of print media can be easily known, making it easy for a user to identify the type of print medium.

[0072] (10) Another aspect is a method for identifying a type of print medium to be printed. This print medium identification method may prepare a trained model for a plurality of types of print media, which is generated by machine learning using the identifier as a teacher, based on the spectral information of the unprinted area of ​​the print medium and an identifier indicating the type of the print medium, acquire spectral information of the unprinted area of ​​the print medium to be printed, and output auxiliary information regarding the type of the print medium using a recognition result obtained by applying the acquired spectral information to the trained model, and accept identification of the type of the print medium using the output auxiliary information. In this way, auxiliary information regarding the type of the print medium is output by applying the spectral information of the unprinted area of ​​the print medium to the trained model, so that the user can easily identify the type of the print medium. In this case, printing may be performed under printing conditions corresponding to the type of the print medium after the print medium is identified.

[0073] (11) Another aspect may be a media information providing device that provides auxiliary information to assist in determining the type of media to be processed in the above configuration, and may include a storage unit that stores a trained model generated by machine learning using a plurality of types of media, which are given spectral information of the background color of the media and an identifier indicating the type of the media, and the identifier is used as a teacher; a spectral information acquisition unit that acquires the spectral information of the background color of the media to be processed; and an output unit that uses the result of applying the acquired spectral information to the trained model to output auxiliary information regarding the type of the media. In this way, by applying the spectral information of the background color of the media to the trained model, auxiliary information regarding the type of media is output, allowing the user to identify the type of media. Furthermore, it is also possible to perform processing under conditions corresponding to the type of media upon receiving this identification.

[0074] (12) In each of the above embodiments, a part of the configuration realized by hardware may be replaced by software. At least a part of the configuration realized by software may be realized by a discrete circuit configuration. In addition, when a part or all of the functions of the present disclosure are realized by software, the software (computer program) may be provided in a form stored in a computer-readable recording medium. The term "computer-readable recording medium" is not limited to portable recording media such as flexible disks and CD-ROMs, but also includes internal storage devices in a computer such as various RAMs and ROMs, and external storage devices fixed to a computer such as a hard disk. In other words, the term "computer-readable recording medium" has a broad meaning including any recording medium to which a data packet can be fixed, not temporarily.

[0075] The present disclosure is not limited to the above-mentioned embodiment, and can be realized in various configurations without departing from the spirit of the present disclosure. For example, the technical features in the embodiments corresponding to the technical features in each aspect described in the Summary of the Invention column can be appropriately replaced or combined to solve some or all of the above-mentioned problems or to achieve some or all of the above-mentioned effects. Furthermore, if the technical feature is not described as essential in this specification, it can be appropriately deleted. [Explanation of symbols]

[0076] 10...printing system, 20...printer, 21...processor, 22...printing control unit, 23...communication unit, 24...display, 25...head transport device, 26...paper transport device, 27...print head, 28...carriage, 30...non-volatile memory, 31...trained model, 33...printing profile, 41...ink cartridge, 43...spectroreflectometer, 50...machine learning device, 51...learning control unit, 53...communication unit, 55...bus, 60...machine learning unit, 61...training model, 63...teacher data, 65...trained model, 70...non-volatile memory, 73...printing profile, 90,91,97,98...selection dialog box, 95...table, 100...media information providing device, 110...control unit, 124...output unit, 126...transport unit, 130...media recognition unit, 131...trained model, 133...processing profile, 143...spectral information acquisition unit, 180...media processing unit

Claims

1. A printing system for printing on a print medium, comprising: A storage unit that stores a trained model that is generated in advance for a plurality of types of printing media by machine learning using an identifier as a teacher, based on spectral information of an unprinted area of ​​the printing medium and an identifier indicating the type of the printing medium; a spectral information acquisition unit that acquires spectral information of an unprinted area of ​​a printing medium on which printing is to be performed; an output unit that outputs auxiliary information regarding the type of the printing medium using a result of applying the acquired spectral information to the stored trained model; an identification unit that accepts identification of the type of the print medium using the auxiliary information output by the output unit; Equipped with a printing condition setting unit that sets printing conditions suitable for the printing medium when the type of the printing medium is identified based on the auxiliary information output by the output unit; When the print medium is not the learned print medium, the printing condition setting unit allows a user to select whether to perform printing using the printing conditions for the learned print medium or to create new printing condition settings and perform printing. Printing system.

2. the spectral information is at least one of a spectral reflectance and a spectral transmittance, the spectral information acquisition unit includes a measurement unit that measures at least one of a spectral reflectance and a spectral transmittance of the unprinted area of ​​the print medium; The printing system of claim 1 .

3. The storage unit is A re-learning unit that re-learns the learned model for a print medium that is determined to be an unknown print medium by the learned model, using spectral information of the unknown print medium and an identifier assigned to the unknown print medium; An update unit that updates the stored trained model using a learning result by the re-learning unit; The printing system according to claim 1 or 2, further comprising:

4. 4. The printing system according to claim 3, further comprising a printing condition setting unit that sets printing conditions suitable for the printing medium determined to be an unknown printing medium.

5. the output unit includes a display unit that displays the auxiliary information, the printing condition setting unit includes an identification unit that allows a user to identify the print medium by using the auxiliary information displayed on the display unit; The printing system of claim 1 .

6. The printing system of claim 5, wherein the output unit displays, on the display unit, information indicating the degree to which the printing medium, about which the spectral information has been acquired by the spectral information acquisition unit, is similar to a trained printing medium that was trained when generating the trained model, as one piece of auxiliary information.

7. The printing system of claim 1, wherein the output unit displays the degree of similarity calculated from the spectral information on the display unit as one piece of auxiliary information indicating the degree to which the printing medium determined to be unlearned is similar to the learned printing medium, in order to create the new printing condition settings.

8. A method for identifying a print medium to be printed, comprising: preparing a trained model for a plurality of types of print media, the trained model being generated by machine learning using the identifier as a teacher, based on the spectral information of an unprinted area of ​​the print medium and an identifier indicating the type of the print medium; Acquire spectral information of an unprinted area of ​​a printing medium to be printed; outputting auxiliary information regarding the type of the printing medium using a recognition result obtained by applying the acquired spectral information to the trained model; Accepting the identification of the type of the print medium using the output auxiliary information; When the type of the print medium is identified based on the output auxiliary information, printing conditions suitable for the print medium are set; If the print medium is not the learned print medium, the user is prompted to select whether to use the print conditions for the learned print medium or to create and use new print conditions when printing. Print media identification method.

9. A media information providing device that provides auxiliary information for assisting in identifying a type of media to be processed, comprising: A storage unit that stores a trained model generated by machine learning using a plurality of types of media, the trained model being given spectral information of the background color of the media and an identifier indicating the type of the media, and the identifier being used as a teacher; a spectral information acquisition unit for acquiring spectral information of a background color of a medium to be processed; An output unit that outputs auxiliary information regarding the type of the media by using a result of applying the acquired spectral information to the trained model; Equipped with a printing condition setting unit that sets printing conditions suitable for the medium when the type of the medium is identified based on the auxiliary information output by the output unit; When the medium is not the learned medium, the printing condition setting unit allows a user to select whether to use the printing conditions for the learned medium or to create new printing condition settings and use the new printing condition settings. Media information providing device.

Images