Printing data generating device
Through machine learning to predict printing speed and adjust RIP setting conditions, the printing speed optimization problem in the prior art is solved and more efficient printing processing is achieved.
Patent Information
- Application Number
- CN202010466481.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2019-05-29
- Filing Date
- 2020-05-28
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2040-11-26
AI Technical Summary
The prior art is difficult to determine the optimal RIP setting conditions for maximizing printing speed, and the printing speed is affected by the complex combination of the original data size and printing setting conditions, making it difficult to optimize.
Using machine learning components, the printing speed is predicted by learning the model, and the RIP setting conditions are adjusted according to the prediction results, such as the number of CPU threads and the number of strips divided by the original data, to ensure that the printing speed is close to or exceeds the processing speed of the printing unit.
It realizes automatic optimization of RIP setting conditions based on specific printing job data and setting conditions, improves printing speed, reduces user burden and improves processing efficiency.
Smart Images

Figure CN112015355B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a print data generating device that generates print data by performing RIP (Raster Image Processor) processing on manuscript data included in a print job. Background Art
[0002] Conventionally, there is known a printing apparatus that receives a print job in a PDL (Page Description Language) format output from a computer and performs printing based on the print job.
[0003] In such a printing apparatus, a received print job in the PDL format is subjected to RIP processing to generate print data (see, for example, Patent Document 1).
[0004] The RIP process is a process of rasterizing a print job after analyzing the text structure of the print job to generate print data. The printing device performs printing based on the print data generated by the RIP process.
[0005] Prior art literature
[0006] Patent Literature
[0007] Patent Document 1: Japanese Patent Application Publication No. 2013-43294 Summary of the invention
[0008] Problem that the invention aims to solve
[0009] Here, the above-mentioned RIP processing is performed based on the prescribed RIP setting conditions. Examples of the RIP setting conditions include the setting conditions of MTR (Multi thread rendering), the number of divided bands of the manuscript data when performing RIP processing, and the number of memories used in the RIP processing. The setting conditions of MTR refer to the number of CPU threads used in the RIP processing. In addition, the number of divided bands of the manuscript data refers to the number of bands to be divided when a sheet of manuscript data is divided into a plurality of bands (areas) and RIP processing is performed on each band.
[0010] Such RIP setting conditions are preferably set to conditions that maximize the printing speed, but the RIP setting conditions that maximize the printing speed vary depending on the data size of the manuscript data contained in the print job and the print setting conditions. In addition, the printing speed is the number of printed sheets per minute (ppm) calculated based on the time from the time the printing device receives the print job until the printed material is discharged. The printing speed is determined based on the data size of the manuscript data, the print setting conditions, and the RIP setting conditions, and the printing speed changes according to these conditions. In addition, maximizing the printing speed here means making the printing speed as close as possible to the processing speed of the printing unit that performs the printing job. The processing speed of the printing unit is the maximum speed (ppm) determined by the mechanical structure of the printing unit, which is a fixed value.
[0011] Furthermore, the data size and printing setting conditions of the manuscript data of the print job can be obtained before printing. However, since there are countless combinations of data size and printing setting conditions, it is difficult to create conditional expressions for these combinations to determine the optimal RIP setting conditions that maximize the printing speed.
[0012] In view of the above circumstances, an object of the present invention is to provide a print data generating apparatus capable of determining optimized RIP setting conditions that maximize the print speed.
[0013] Solutions for solving problems
[0014] The printing data generating device of the present invention comprises: a machine learning unit, which has a learning completion model, wherein the learning completion model is a model obtained by machine learning using the relationship between the data size of manuscript data of multiple printing jobs, printing setting conditions and RIP (Raster Image Processor) setting conditions and the printing speed when the manuscript data is printed by the printing unit as input; a printing speed prediction unit, which predicts the printing speed of the printing job of the printing object by inputting the data size of the manuscript data of the printing job of the printing object, the printing setting conditions and any RIP setting conditions into the learning completion model; a RIP setting condition determination unit, which determines the optimized RIP setting conditions that make the printing speed predicted by the printing speed prediction unit close to the processing speed of the printing unit or above the processing speed; and a RIP processing unit, which performs RIP (Raster Image Processor) processing of the printing job of the printing object based on the optimized RIP setting conditions determined by the RIP setting condition determination unit and generates printing data, and outputs the printing data to the printing unit.
[0015] Effects of the Invention
[0016] According to the printing data generating device of the present invention, there is a machine learning unit, which has a learning completion model. The learning completion model is a model obtained by machine learning based on the relationship between the data size of the manuscript data of multiple printing jobs, printing setting conditions and RIP setting conditions and the printing speed when the manuscript data is printed by the printing unit as input. The printing speed of the printing job of the printing object is predicted by inputting the data size of the manuscript data of the printing job of the printing object, the printing setting conditions and any RIP setting conditions into the learning completion model, and the optimized RIP setting conditions are determined to make the predicted printing speed close to the processing speed of the printing unit or above the processing speed, so that the optimized RIP setting conditions that maximize the actual printing speed can be determined. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Figure 1 This is a block diagram showing a schematic configuration of a printing apparatus using a first embodiment of the print data generating apparatus of the present invention.
[0018] Figure 2 This is a diagram showing a configuration example of hardware of a printing apparatus using the first embodiment of the print data generating apparatus of the present invention.
[0019] Figure 3 This is a graph showing the relationship between the number of threads of the RIP process and productivity (ppm) in the case of the color mode (automatic color printing).
[0020] Figure 4 This is a graph showing the relationship between the number of threads of the RIP process and productivity (ppm) in the case of the color mode (grayscale).
[0021] Figure 5 This is a flowchart for explaining the processing of the printing apparatus according to the first embodiment.
[0022] Figure 6 This is a diagram showing an example of a table in which predicted printing speeds corresponding to various RIP setting conditions are preset.
[0023] Figure 7 This is a block diagram showing a schematic configuration of a printing apparatus using a second embodiment of the print data generating apparatus of the present invention.
[0024] Figure 8 This is a block diagram showing a schematic configuration of a printing apparatus according to a third embodiment using the print data generating apparatus of the present invention.
[0025] Description of Reference Numerals
[0026] 1, 2, 3: printing device; 10: printing data generating unit; 11: prepress inspection unit; 12: machine learning unit; 13: printing speed prediction unit; 14: RIP setting condition determination unit; 15: RIP processing unit; 16: teacher data acquisition unit; 20: printing unit; 30: input acceptance unit; 40: printing data generating unit; 41: prepress inspection unit; 42: machine learning unit; 43: RIP setting condition determination unit; 44: RIP processing unit; 50: printing data generating unit; 51: prepress inspection unit; 52: first machine learning unit; 53: printing speed prediction unit; 54: RIP setting condition determination unit; 55: RIP processing unit; 56: teacher data acquisition unit; 57: second machine learning unit. DETAILED DESCRIPTION
[0027] Hereinafter, a printing apparatus according to a first embodiment using a print data generating apparatus according to the present invention will be described in detail with reference to the drawings. Figure 1 This is a block diagram showing a schematic configuration of the printing device 1 according to the present embodiment.
[0028] like Figure 1 As shown, the printing device 1 of this embodiment includes a print data generating unit 10, a printing unit 20, and an input receiving unit 30. In this embodiment, the print data generating unit 10 corresponds to the print data generating device of the present invention.
[0029] The print data generating section 10 of the present embodiment includes a preflight section 11 , a machine learning section 12 , a print speed prediction section 13 , a RIP setting condition determination section 14 , a RIP processing section 15 , and a teacher data acquisition section 16 .
[0030] in addition, Figure 2 1 is a diagram showing a configuration example of hardware of the printing device 1 according to the present embodiment. Figure 2 As shown in FIG. 1 , the print data generating unit 10 of the printing device 1 of the present embodiment includes a CPU (Central Processing Unit), a semiconductor memory such as a ROM (Read Only Memory) and a RAM (Random Access Memory), a storage such as a hard disk, and a communication I / F. Figure 2 Although only one ROM and one RAM are shown in FIG. 1 , the print data generating unit 10 of the present embodiment is provided with a plurality of ROMs and RAMs.
[0031] A print data generation program is installed in the semiconductor memory or the hard disk of the print data generation unit 10 , and each unit of the print data generation unit 10 functions when the CPU executes the print data generation program.
[0032] In addition, a control program for controlling the operation of the printing section 20 and the input receiving section 30 is installed in the semiconductor memory or the hard disk of the print data generating section 10, and the CPU executes the control program to control the operation of the printing section 20 and the input receiving section 30. That is, the CPU in the print data generating section 10 makes each section of the print data generating section 10 function and controls the operation of the printing section 20 and the input receiving section 30.
[0033] Furthermore, in the present embodiment, the functions of the various parts of the printing data generating unit 10 are implemented by a printing data generating program, and the control of the printing unit 20 and the input receiving unit 30 is implemented by a control program, but it is not limited thereto, and part or all of the functions or controls may also be implemented by hardware such as ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), or other circuits.
[0034] return Figure 1 The pre-press inspection unit 11 receives a print job including manuscript data output from a computer via a communication I / F, for example, and analyzes the print job. The pre-press inspection unit 11 of this embodiment performs pre-press inspection of the print job, obtains the data size and print setting conditions of the manuscript data included in the print job, and outputs the data size and print setting conditions to the print speed prediction unit 13 and the teacher data acquisition unit 16. In addition, the pre-press inspection unit 11 is not limited to receiving the print job output from the computer, and may also receive a print job including manuscript data read by a scanner or the like.
[0035] When a print job includes document data of a plurality of pages, the data size of the document data is the total value of the data sizes of the document data of the plurality of pages.
[0036] Printing setting conditions include, for example, color printing or monochrome printing, single-sided printing or double-sided printing, and N-up printing, but are not limited to these. As long as the setting conditions are related to printing, other conditions may also be included.
[0037] The machine learning unit 12 has a learned model obtained by performing machine learning using as input the relationship between the data size of manuscript data of a plurality of printing jobs, printing setting conditions, RIP setting conditions, and the printing speed when the printing unit 20 prints the manuscript data.
[0038] The learned model may be generated based on a print job input to the printing device 1 , but may also be generated based on print results of a print job in another printing device having a printing unit equivalent to the printing device 1 and stored in the machine learning unit 12 .
[0039] The RIP setting conditions are conditions used when performing RIP processing in the RIP processing unit 15, such as the setting conditions of MTR, the number of divided strips of the original data when performing RIP processing, and the number of memories used in the RIP processing, but as long as they are conditions used in the RIP processing, other conditions may also be included.
[0040] The setting condition of MTR refers to the number of CPU threads used in RIP processing.
[0041] Here, the more threads used in the RIP process, the faster the RIP process itself can be performed, but it cannot be said that the more threads used in the RIP process, the higher the productivity. In addition, productivity is synonymous with printing speed, which is the number of printed sheets per minute (ppm) calculated based on the time from the printing device 1 receiving the print job to the time when the printed material is discharged.
[0042] This is because the CPU is also used in processes other than RIP processing. When a large number of threads are allocated to RIP processing, the number of CPU threads used in other processes decreases, thereby reducing processing efficiency and increasing processing time for the entire process.
[0043] Furthermore, the optimum number of threads for RIP processing to improve productivity is related to the print setting conditions as described above. Figure 3 This is a graph showing the relationship between the number of threads in the RIP process and productivity (ppm) in the case of the color mode (automatic color printing). Figure 4 : is a graph showing the relationship between the number of threads of RIP processing and productivity (ppm) in the case of color mode (grayscale). Figure 3 As shown, in the case of color mode (automatic color printing), productivity improves when the number of threads reaches 3, and productivity is almost unchanged between the number of threads 3 and the number of threads 4. That is, in this case, it is better to set the number of threads for RIP processing to 3 and allocate the number of CPU threads to processes other than RIP processing.
[0044] In addition, if Figure 4 As shown, in the case of the color mode (grayscale), the greater the number of threads assigned to the RIP process, the lower the productivity.
[0045] As described above, the combination of the number of threads used in the RIP process and the print setting conditions is related to productivity, that is, the print speed of the printing unit 20 .
[0046] In addition, the number of divided strips of manuscript data in the RIP setting condition refers to the number of strips to be divided when a sheet of manuscript data is divided into a plurality of strips (areas) and RIP processing is performed on each strip. When the number of divided strips is small, the data size of the manuscript data to be RIP processed at one time becomes larger, so it is necessary to increase the number of memories (the number of semiconductor memories) used in the RIP processing described later. As a result, the number of memories allocated for processing other than RIP processing becomes smaller, and the processing becomes slower.
[0047] On the other hand, when the number of divided bands is large, the data size of the document data to be RIP processed at one time becomes smaller, so the amount of memory used in the RIP process can be reduced, but the RIP process must be repeatedly performed for each band, so the RIP process becomes slower accordingly.
[0048] As described above, the number of divided bands of document data and the number of memories used in the RIP process in the RIP setting conditions are also related to productivity, that is, the printing speed of the printing unit 20 .
[0049] The learning model of the machine learning unit 12 is a model obtained by performing machine learning using as input the data size of the manuscript data of the printing job considered to be related to the printing speed of the printing unit 20, the printing setting conditions, and the relationship between the above-mentioned RIP setting conditions and the actual printing speed at that time.
[0050] The printing speed prediction unit 13 predicts the printing speed of the printing target job by inputting the data size of the manuscript data of the printing target job, the printing setting conditions, and any RIP setting conditions to the learned model stored in the machine learning unit 12 .
[0051] Specifically, the printing speed prediction unit 13 of the present embodiment obtains the data size and printing setting conditions of the manuscript data of the print job of the print target output from the prepress inspection unit 11. Then, the printing speed prediction unit 13 first inputs the acquired data size and printing setting conditions of the manuscript data and the preset initial setting RIP setting conditions into the learning model stored in the machine learning unit 12.
[0052] Then, the printing speed prediction unit 13 acquires the printing speed output from the learned model stored in the machine learning unit 12 , thereby predicting the printing speed of the print job to be printed.
[0053] The RIP setting condition determination unit 14 determines the optimal RIP setting condition that brings the printing speed predicted by the printing speed prediction unit 13 close to the processing speed of the printing unit 20 .
[0054] Specifically, the RIP setting condition determination unit 14 of the present embodiment first compares the printing speed predicted by the printing speed prediction unit 13 based on the RIP setting conditions set initially with the preset processing speed of the printing unit 20. The processing speed of the printing unit 20 refers to the processing speed (ppm) determined by the mechanical structure of the printing unit 20. The processing speed of the printing unit 20 is set and input by the user in the input receiving unit 30 and stored in the RIP setting condition determination unit 14, for example.
[0055] Then, when the predicted printing speed is slower than the processing speed of the printing unit 20, that is, when the processing capacity of the printing unit 20 cannot be fully utilized, the RIP setting condition determination unit 14 changes the RIP setting conditions initially set, and outputs the changed RIP setting conditions to the printing speed prediction unit 13. The printing speed prediction unit 13 inputs the input changed RIP setting conditions, the data size of the manuscript data of the print job to be printed, and the printing setting conditions to the learning model, and predicts the printing speed again.
[0056] The printing speed predicted by the printing speed prediction unit 13 is input again to the RIP setting condition determination unit 14. The RIP setting condition determination unit 14 determines whether the re-predicted printing speed is within a preset threshold range based on the processing speed of the printing unit 20. If it is not within the threshold range, it is determined that the predicted printing speed is not close to the processing speed of the printing unit 20, and the RIP setting condition is changed again, and the changed RIP setting condition is output to the printing speed prediction unit 13.
[0057] As described above, the RIP setting condition determination unit 14 determines whether the printing speed predicted by the printing speed prediction unit 13 is within the preset threshold range, and if it is not within the threshold range, the RIP setting conditions are sequentially changed and output to the printing speed prediction unit 13. Then, the printing speed prediction unit 13 sequentially predicts the printing speed using the learned model based on the input of the changed RIP setting conditions, and sequentially outputs the predicted printing speed to the RIP setting condition determination unit 14.
[0058] Then, by repeatedly changing the RIP setting conditions and predicting the printing speed, the RIP setting condition determination unit 14 determines the optimized RIP setting conditions that make the printing speed predicted by the printing speed prediction unit 13 fall within a preset threshold range, and outputs the optimized RIP setting conditions to the RIP processing unit 15.
[0059] The RIP processing unit 15 performs RIP processing on the print job to be printed based on the optimized RIP setting conditions determined by the RIP setting condition determination unit 14 , generates print data, and outputs the print data to the printing unit 20 .
[0060] When the printing of the print job of the print target is completed in the printing unit 20 based on the print data generated by the RIP processing unit 15, the teacher data acquisition unit 16 acquires the printing speed during the printing, the data size of the manuscript data of the print job of the print target, the printing setting conditions, and the above-mentioned optimized RIP setting conditions as teacher data, and outputs the teacher data to the machine learning unit 12. The machine learning unit 12 further performs machine learning of the learning completion model using the input teacher data. By performing machine learning of the learning completion model in this way, the prediction accuracy of the printing speed can be improved.
[0061] The printing unit 20 performs printing on a printing medium such as paper or film using the printing data output from the RIP processing unit 15. The printing unit 20 of the present embodiment can be, for example, a printing engine having an inkjet head that ejects ink onto a printing medium to perform printing, but is not limited to such an inkjet printing engine, and may be a laser printing engine that transfers toner to a printing medium using a photosensitive drum to perform printing, or may be a stencil printing engine.
[0062] The input receiving unit 30 is composed of, for example, a touch panel, and displays various information and receives various setting inputs from the user. The input receiving unit 30 displays, for example, a print condition setting input screen, and receives the setting input of the print setting conditions from the user on the print condition setting input screen. The print setting conditions set and input on the print condition setting input screen are included in the print job in a manner associated with the document data read by a scanner or the like, for example.
[0063] Next, refer to Figure 5 The processing of the printing device 1 according to the present embodiment will be described with reference to the flowchart shown.
[0064] First, the prepress check section 11 receives a print job output from a computer or the like (S10). The prepress check section 11 performs prepress check of the print job as described above, and acquires the number of pages of manuscript data included in the print job, the data size of the manuscript data, and print setting conditions (S12).
[0065] Then, the number of pages of the manuscript data, the data size of the manuscript data, and the printing setting conditions output from the pre-press inspection unit 11 are input to the printing speed prediction unit 13. The printing speed prediction unit 13 compares the input number of pages with a preset threshold value (S14), and when the input number of pages exceeds the threshold value (S14: "Yes"), the printing speed prediction unit 13 predicts the printing speed of the printing job using the data size of the input manuscript data, the printing setting conditions, and the initially set RIP setting conditions (S16). Specifically, the printing speed prediction unit 13 predicts the printing speed by inputting the data size of the manuscript data, the printing setting conditions, and the initially set RIP setting conditions to the learning model stored in the machine learning unit 12 as described above. In addition, as the above threshold, for example, one page is set, but it is not limited to this, and other numbers of pages can also be set.
[0066] The printing speed predicted by the printing speed prediction unit 13 is input to the RIP setting condition determination unit 14 , and the RIP setting condition determination unit 14 compares the input printing speed with the processing speed of the printing unit 20 set in advance ( S18 ).
[0067] Then, when the predicted printing speed is slower than the processing speed of the printing unit 20 ( S18 : Yes), the RIP setting condition determination unit 14 changes the above-mentioned initial RIP setting conditions ( S20 ) and outputs the changed RIP setting conditions to the printing speed prediction unit 13 .
[0068] The print speed prediction unit 13 inputs the input changed RIP setting conditions, the data size of the document data of the print job to be printed, and the print setting conditions into the learned model, and predicts the print speed again ( S22 ).
[0069] The printing speed predicted again in the printing speed prediction unit 13 is input again to the RIP setting condition determination unit 14. The RIP setting condition determination unit 14 determines whether the printing speed predicted again is within the preset threshold range (S24). When the printing speed predicted again is not within the preset threshold range (S24: "No"), the RIP setting condition determination unit 14 changes the RIP setting condition again (S20), and outputs the changed RIP setting condition to the printing speed prediction unit 13. Then, the printing speed prediction unit 13 predicts the printing speed again based on the input of the changed RIP setting condition and using the learning completed model, and outputs it to the RIP setting condition determination unit 14 (S22).
[0070] Then, the RIP setting condition determination unit 14 changes the RIP setting condition and the printing speed prediction unit 13 predicts the printing speed repeatedly until the printing speed predicted by the printing speed prediction unit 13 enters the preset threshold range. Then, the RIP setting condition determination unit 14 determines the RIP setting condition at the time point when the predicted printing speed enters the preset threshold range as the optimized RIP setting condition, and outputs the optimized RIP setting condition to the RIP processing unit 15 (S24: Yes, S26).
[0071] Next, the RIP processing unit 15 performs RIP processing on the print job to be printed based on the optimized RIP setting conditions determined by the RIP setting condition determination unit 14 to generate print data, and outputs the print data to the printing unit 20 ( S28 ).
[0072] On the other hand, when it is determined in S14 that the number of pages input is below the threshold value (S14: "No"), the prediction process of the printing speed by the printing speed prediction unit 13 and the determination process of the optimized RIP setting conditions by the RIP setting condition determination unit 14 are not performed. When the number of pages in a print job is small, the RIP processing itself does not require much time, and sometimes the above-mentioned printing speed prediction process and the determination process of the optimized RIP setting conditions will cause the processing time to be longer. Therefore, when the number of pages in a print job is small, the processing time can be shortened by omitting the printing speed prediction process and the determination process of the optimized RIP setting conditions. In addition, the burden on the CPU can be reduced, thereby improving the processing efficiency.
[0073] Then, in this case, the RIP processing unit 15 performs RIP processing based on the initially set RIP setting conditions to generate print data ( S28 ), and outputs the print data to the printing unit 20 .
[0074] In addition, if it is determined in S18 that the printing speed predicted based on the initially set RIP setting conditions is greater than the processing speed of the printing unit 20 (S18: No), the RIP setting condition determination unit 14 does not perform the determination process of the optimized RIP setting conditions. In this way, by omitting the determination process of the optimized RIP setting conditions when it is not necessary, the processing time can be shortened. In addition, the burden on the CPU can be reduced, thereby improving the processing efficiency.
[0075] Then, in this case as well, the RIP processing unit 15 performs RIP processing based on the initially set RIP setting conditions to generate print data ( S28 ), and outputs the print data to the printing unit 20 .
[0076] Then, the printing unit 20 performs printing on the printing medium based on the input printing data ( S30 ).
[0077] Then, when the printing of the print job in the printing unit 20 is completed, the printing speed during the printing, the data size of the manuscript data of the print job, the printing setting conditions, and the above-mentioned optimized RIP setting conditions are obtained by the teacher data acquisition unit 16 as teacher data (S32). The teacher data acquisition unit 16 outputs the teacher data to the machine learning unit 12. The machine learning unit 12 further performs machine learning of the learning completion model using the input teacher data (S34).
[0078] According to the printing device 1 of the first embodiment, the data size of the manuscript data of the print job to be printed, the print setting conditions, and any RIP setting conditions are input to the learning model of the machine learning unit 12, so as to predict the print speed of the print job to be printed, and determine the optimal RIP setting conditions that make the predicted print speed close to the processing speed of the printing unit, so that the optimal RIP setting conditions that make the actual print speed the maximum can be determined. In addition, the burden on the user is reduced by improving the print speed.
[0079] In addition, in the printing device 1 of the above-mentioned first embodiment, the RIP setting condition determination unit 14 determines the RIP setting condition at the time point when the predicted printing speed enters the pre-set threshold range as the optimized RIP setting condition, but is not limited to this. The RIP setting condition at the time point when the predicted printing speed becomes higher than the processing speed of the printing unit 20 can also be determined as the optimized RIP setting condition.
[0080] In addition, in the printing device 1 of the first embodiment described above, the RIP setting condition is changed by the RIP setting condition determination unit 14 to repeatedly predict the printing speed, thereby determining the optimal RIP setting condition. However, when the printing speed is predicted each time the RIP setting condition is changed, the calculation time may be longer.
[0081] Therefore, the above-mentioned learning completion model can also be used to pre-calculate the printing speed corresponding to various RIP setting conditions and store them in the form of a table in the RIP setting condition determination unit 14 in advance. The RIP setting condition determination unit 14 determines the optimal RIP setting conditions by referring to the above-mentioned table based on the data size and printing setting conditions obtained by the prepress inspection unit 11, and the processing speed of the printing unit 20.
[0082] Figure 6 It is a figure which shows typically an example of the said table. Figure 6The predicted printing speed in the table shown is a value obtained using the above-mentioned learning model. Specifically, for example, when the data size of the manuscript data is data size 1 and the printing setting condition is printing setting condition 1, the printing speed predicted by the learning model by changing the RIP setting condition to RIP setting conditions 1 to 4 is predicted printing speed 1 to 4. In addition, data size 1 to data size 4 are set in the range of data size such as 1 kilobyte to 50 kilobytes, 51 kilobytes to 100 kilobytes, 101 kilobytes to 150 kilobytes, etc., respectively.
[0083] For example, when the data size of the print job to be printed is within the range of data size 1 and the print setting condition is print setting condition 1, the RIP setting condition determination unit 14 refers to the processing speed of the printing unit 20. Figure 6 The predicted printing speed closest to the processing speed of the printing unit 20 is determined from the predicted printing speeds 1 to 4 by using the table shown in the figure. Then, when the predicted printing speed closest to the processing speed of the printing unit 20 is the predicted printing speed 1, the RIP setting condition determination unit 14 determines the RIP setting condition 1 corresponding to the predicted printing speed 1 as the optimal RIP setting condition.
[0084] By determining the optimal RIP setting condition using the preset table as described above, it is possible to shorten the processing time for determining the optimal RIP setting condition.
[0085] In addition, in the printing device 1 of the above-mentioned first embodiment, the data size of the original data of the printing job, the printing setting conditions and the RIP setting conditions are used as input information of the learning completion model, but other information can also be added as input information to improve the prediction accuracy of the printing speed output by the learning completion model.
[0086] Specifically, for example, when the user is performing certain setting input operations in the input receiving unit 30, the CPU in the print data generating unit 10 controls the operation of the input receiving unit 30, so the CPU occupancy rate of the RIP processing is reduced, which may affect the printing speed (productivity).
[0087] In addition, the operation control is not limited to the input acceptance unit 30. For example, when the printing unit 20 is printing a previous printing job, the operation control of the printing unit 20 is performed by the CPU in the printing data generation unit 10. Therefore, the CPU occupancy rate of the RIP processing is still reduced, which may affect the printing speed (productivity).
[0088] In addition, for example, when the inkjet head is cleaned or consumables such as ink, toner or printing paper are replaced during the RIP processing of a printing job, the printing operation of the printing unit 20 is stopped during this period, and the CPU occupancy of the RIP processing increases, which may affect the printing speed (productivity).
[0089] That is, the current operating state of the printing device 1, such as the operating state of the input receiving unit 30 and the operating state of the printing unit 20, or the operating state of the printing unit 20 during the cleaning operation and the replacement of consumables, etc., or the operating state of the printing device 1 predicted during the processing of the print job of the print target, such as the operating state of the printing unit 20 during the cleaning operation and the replacement of consumables, may be used as additional input information of the learning model. In this way, the prediction accuracy of the printing speed using the learning model can be improved.
[0090] The above-mentioned additional input information is acquired by the CPU and used as input information when generating a learned model and when predicting a printing speed using the generated learned model.
[0091] In addition, regarding the information on the operation state of the printing device 1 predicted during the processing of the printing job, for example, in the case of a cleaning operation, since the cleaning operation is performed according to the preset number of printed sheets, it is possible to determine whether the cleaning operation is performed during the processing of the printing job to be printed (whether the operation of the printing unit 20 has stopped) by obtaining the number of printed sheets since the previous cleaning operation. In addition, regarding the replacement timing of consumables, it is possible to determine whether the replacement timing of consumables has occurred during the processing of the printing job to be printed (whether the operation of the printing unit 20 has stopped) by obtaining the remaining amount of consumables. And, these determination results can be used as input information for the learning completion model.
[0092] In addition, in the printing device 1 of the first embodiment described above, the data size of the manuscript data of the printing job is used as input information of the learning completion model, but the speed of the RIP processing of the manuscript data also changes according to the resolution of the manuscript data, the ratio of photos to text, or the printing rate, etc., which affects the printing speed (productivity). Therefore, the input information of the learning completion model can also be set to include feature quantities of the manuscript data such as the resolution of the manuscript data, the ratio of photos to text, or the printing rate. In this way, the prediction accuracy of the printing speed using the learning completion model can be improved. In addition, the feature quantities of the manuscript data mentioned above are also obtained and used as input information when generating the learning completion model and when predicting the printing speed using the generated learning completion model.
[0093] Next, a printing device 2 according to a second embodiment using the print data generating device of the present invention is described. In the printing device 1 according to the first embodiment, the printing speed is predicted based on the data size of the manuscript data of the print job, the print setting conditions, and the RIP setting conditions using the learned model, and the optimized RIP setting conditions that make the predicted printing speed close to the processing speed of the printing unit 20 are obtained by changing the RIP setting conditions. However, if this method is implemented for a plurality of print jobs, a combination of the data size of the manuscript data and the print setting conditions of a plurality of predetermined print jobs and the optimized RIP setting conditions for the print job can be obtained.
[0094] The printing device 2 of the second embodiment has a learning completion model newly generated using the data size and printing setting conditions of the manuscript data of the above-mentioned multiple printing jobs and the group of the optimized RIP setting conditions. Specifically, the learning completion model of the second embodiment is a learning completion model obtained by machine learning using the relationship between the data size and printing setting conditions of the manuscript data of each group and the optimized RIP setting conditions of each group as input. In addition, the learning completion model in the printing device 1 of the first embodiment is equivalent to the first learning completion model of the present invention, and the learning completion model newly added in the printing device 2 of the second embodiment is equivalent to the second learning completion model of the present invention.
[0095] The printing device 2 of the second embodiment inputs the data size of the manuscript data of the print job to be printed and the print setting conditions to the newly generated learned model (second learned model), thereby determining the optimized RIP setting conditions for the print job to be printed.
[0096] Figure 7 2 is a block diagram showing a schematic structure of a printing device 2 according to a second embodiment. Figure 7 As shown, the printing device 2 according to the second embodiment includes a print data generating unit 40 , a printing unit 20 , and an input accepting unit 30 .
[0097] The print data generating section 40 of the second embodiment includes a pre-press inspection section 41 , a machine learning section 42 , a RIP setting condition determining section 43 , and a RIP processing section 44 .
[0098] The prepress inspection unit 41 receives a print job including manuscript data output from a computer or the like, performs prepress inspection of the print job, and obtains the data size and print setting conditions of the manuscript data included in the print job, similarly to the prepress inspection unit 11 of the first embodiment, and outputs the data size and print setting conditions to the RIP setting condition determination unit 43. In addition, the data size and print setting conditions of the manuscript data are the same as those of the first embodiment.
[0099] The RIP setting condition determination unit 43 obtains the data size and printing setting conditions of the manuscript data output from the prepress inspection unit 41, and inputs the obtained data size and printing setting conditions of the manuscript data into the learned model (second learned model) pre-stored in the machine learning unit 42. Then, the RIP setting condition determination unit 43 determines the optimized RIP setting conditions by obtaining the optimized RIP setting conditions output from the learned model of the machine learning unit 42.
[0100] The RIP processing unit 44 is the same as the RIP processing unit 15 of the first embodiment.
[0101] The printing unit 20 and the input receiving unit 30 are also the same as those of the printing device 1 according to the first embodiment.
[0102] According to the printing device 2 of the second embodiment, the machine learning unit 42 includes a learning model obtained by machine learning using the data size of the manuscript data of a plurality of printing jobs and the printing setting conditions and the relationship between the optimized RIP setting conditions of the plurality of printing jobs as inputs, and the optimized RIP setting conditions of the printing jobs are determined by inputting the data size of the manuscript data of the printing job to be printed and the printing setting conditions to the learning model of the machine learning unit 42. That is, in the printing device 2 of the second embodiment, after the learning model (second learning model) of the machine learning unit 42 is temporarily completed, the prediction process of the printing speed and the determination process of the optimized RIP setting conditions using the learning model (first learning model) as in the printing device 1 of the first embodiment are not performed, but the optimized RIP setting conditions can be determined only by inputting the data size of the manuscript data and the printing setting conditions. Therefore, the optimized RIP setting conditions that maximize the actual printing speed can be determined by simpler calculation processing than that of the printing device 1 of the first embodiment.
[0103] In addition, in the printing device 2 of the second embodiment, similarly to the first embodiment, the RIP setting condition determination unit 43 may compare the number of pages of the manuscript data output from the pre-press inspection unit 41 with a preset threshold value, and when the number of pages input exceeds the threshold value, the above-mentioned learning completion model is used to perform the determination process of optimizing the RIP setting conditions, and when the number of pages input is less than the threshold value, the determination process of optimizing the RIP setting conditions is not performed. Then, in this case, the RIP processing unit 44 performs RIP processing based on the initially set RIP setting conditions and generates print data, and outputs the print data to the printing unit 20.
[0104] In addition, in the printing device 2 of the second embodiment, the above-mentioned current operation state of the printing device 2, information on the operation state of the printing device 2 predicted in the processing of the print job to be printed, and feature quantities of the manuscript data may be used as additional input information for the learning completion model. This can improve the accuracy of the optimized RIP setting conditions.
[0105] Furthermore, when the additional information is used as input information of the learning model of the second embodiment, the additional information is also used when acquiring a combination of the data size and printing setting conditions of a plurality of manuscript data used when generating the learning model of the second embodiment and the optimized RIP setting conditions. That is, when acquiring the above combination, the additional information is also used as input information of the learning model of the second embodiment.
[0106] Furthermore, the printing device 2 of the second embodiment may include the learned model of the first embodiment (first learned model). Figure 8 This is a block diagram showing a schematic configuration of a printing device 3 according to a third embodiment that includes both the learned model according to the first embodiment and the learned model according to the second embodiment.
[0107] The printing device 3 of the third embodiment includes a print data generating unit 50, a printing unit 20, and an input accepting unit 30. The printing unit 20 and the input accepting unit 30 of the printing device 3 of the third embodiment are the same as those of the first and second embodiments.
[0108] The pre-press inspection unit 51 , the printing speed prediction unit 53 , and the RIP processing unit 55 in the print data generating unit 50 are also the same as those in the first and second embodiments.
[0109] The first machine learning unit 52 and the teacher data acquisition unit 56 are the same as the machine learning unit 12 and the teacher data acquisition unit 16 of the first embodiment, and the second machine learning unit 57 is the same as the machine learning unit 42 of the second embodiment.
[0110] In the printing device 3 of the third embodiment, similarly to the first embodiment, for a plurality of arbitrary printing jobs, the learning completion model of the first machine learning unit 52 is used to perform prediction processing of the printing speed and determination processing of the optimized RIP setting conditions, and the data size of the manuscript data of the plurality of printing jobs and the combination of the printing setting conditions and the optimized RIP setting conditions are input as teacher data to the second machine learning unit 57 for machine learning, and a learning completion model similar to the learning completion model of the second embodiment is generated.
[0111] Then, the RIP setting condition determination unit 54 of the printing device 3 of the third embodiment inputs the data size of the manuscript data of the print job to be printed and the print setting conditions to the learned model of the second machine learning unit 57 to determine the optimal RIP setting conditions for the print job to be printed.
[0112] Then, the RIP process is performed using the determined optimal RIP setting conditions to perform the printing process.
[0113] After the printing process is completed, the data size of the manuscript data of the printing job to be processed and the printing setting conditions are input into the first learning completion model to perform the prediction process of the printing speed, and the above-mentioned determination process of the optimized RIP setting conditions is performed, thereby generating the teacher data consisting of the data size of the manuscript data and the printing conditions and the optimized RIP setting conditions. Then, by inputting the teacher data into the second machine learning unit 57, the machine learning of the learning completion model of the second machine learning unit 57 can be further performed. In this way, the accuracy of the optimized RIP setting conditions can be improved.
[0114] In addition, other operations of the printing device 3 of the third embodiment are the same as those of the first embodiment or the second embodiment.
[0115] In addition, the first to third embodiments described above are methods of installing the print data generation device of the present invention in a printing device, but are not limited thereto, and the print data generation unit of the first to third embodiments described above may also be provided for an external controller for performing various image processing on manuscript data of a printing job. The external controller described above does not include a printing unit, but is provided in a manner of being externally mounted to various printing devices. When the print data generation device of the present invention is applied to an external controller, it is preferred that a learning completion model for each of various printing methods such as an inkjet method, a laser method, and a stencil printing method is provided. Furthermore, it is preferred that the learning completion model be switched for use according to the method of the printing unit connected to the external controller.
[0116] The following additional remarks are also disclosed regarding the print data generating device of the present invention.
[0117] (Note)
[0118] In the printing data generating device of the present invention, when the printing speed predicted by the printing speed prediction unit based on the initially set RIP setting conditions is greater than the processing speed of the printing unit, the RIP setting condition determination unit does not perform the determination processing of the optimized RIP setting conditions, and the RIP processing unit performs RIP processing of the printing job of the printing object based on the initially set RIP setting conditions and generates printing data, and outputs the printing data to the printing unit.
[0119] In addition, in the printing data generating device of the present invention, when the number of pages of the manuscript data of the printing job of the printing object is below a preset threshold value, the RIP setting condition determination unit does not perform the prediction processing of the printing speed and the determination processing of the optimized RIP setting conditions, and the RIP processing unit performs RIP processing of the printing job of the printing object based on the preset RIP setting conditions and generates printing data, and outputs the printing data to the printing unit.
[0120] In addition, the printing data generating device of the present invention is provided with a teacher data acquisition unit. When the printing of the printing job of the printing object is completed in the printing unit, the teacher data acquisition unit acquires the printing speed during the printing, the data size of the original data of the printing job of the printing object, the printing setting conditions and the optimized RIP setting conditions as teacher data, and outputs the teacher data for machine learning of the completion model.
[0121] Another aspect of the present invention provides a printing data generating device comprising: a machine learning unit having a first learning completion model and a second learning completion model, wherein the first learning completion model is a model that combines the data size of the manuscript data of a plurality of printing jobs, the printing setting conditions and the RIP (Raster Image Processing) The invention relates to a model obtained by performing machine learning using the relationship between a setting condition of a processing unit (RIP) and a printing speed when the manuscript data is printed by the printing unit as input, the second learning completion model is a model obtained by performing machine learning using a group of data size of manuscript data of any printing job, printing setting conditions and optimized RIP setting conditions obtained by performing prediction processing and decision processing on a plurality of arbitrary printing jobs using the first learning completion model, and the relationship between the data size and printing setting conditions of each group and the optimized RIP setting conditions of each group as input, the prediction processing predicts the printing speed of any printing job by inputting the data size of manuscript data of any printing job, printing setting conditions and any RIP setting conditions, the decision processing determines the optimized RIP setting conditions that make the predicted printing speed close to the processing speed of the printing unit or above the processing speed; a RIP setting condition determination unit, which uses the second learning completion model to determine the optimized RIP setting conditions for the printing job of the printing object; and a RIP processing unit, which performs RIP (Raster Image Processing) of the printing job of the printing object based on the optimized RIP setting conditions determined by the RIP setting condition determination unit. Processor) processes and generates printing data, and outputs the printing data to the printing unit.
Claims
1. A printing data generating device, comprising: a machine learning unit having a learning model obtained by performing machine learning using as input a relationship between data size of manuscript data of a plurality of printing jobs, printing setting conditions, and RIP setting conditions, and a printing speed when the manuscript data is printed by the printing unit; a printing speed prediction unit that predicts the printing speed of the printing job of the printing object by inputting the data size of the manuscript data of the printing job of the printing object, the printing setting conditions, and any RIP setting conditions to the learned model; a RIP setting condition determination unit for determining an optimized RIP setting condition so that the printing speed predicted by the printing speed prediction unit approaches the processing speed of the printing unit or exceeds the processing speed; as well as a raster image processor processing unit that performs raster image processor processing of the print job of the print object based on the optimized raster image processor setting conditions determined by the raster image processor setting condition determination unit to generate print data, and outputs the print data to the printing unit, The printing speed refers to the number of printed sheets per unit time calculated based on the time from when a print job is received until when printed materials are discharged.
2. The print data generating device according to claim 1, in, When the printing speed predicted by the printing speed prediction unit based on the initially set RIP setting conditions is greater than the processing speed of the printing unit, the RIP setting condition determination unit does not perform the determination process of the optimized RIP setting conditions. The RIP processing unit performs RIP processing of the print job of the print target based on the RIP setting conditions that are initially set to generate print data, and outputs the print data to the printing unit.
3. The print data generating device according to claim 1, in, When the number of pages of the manuscript data of the print job to be printed is less than a preset threshold value, the print speed prediction process and the optimal RIP setting condition determination process are not performed, The RIP processing unit performs RIP processing of the print job of the print target based on a preset RIP setting condition to generate print data, and outputs the print data to the printing unit.
4. The print data generating device according to any one of claims 1 to 3, in, The invention is provided with a teacher data acquisition unit. When the printing of the printing job of the printing object in the printing unit is completed, the teacher data acquisition unit acquires the printing speed during the printing, the data size of the manuscript data of the printing job of the printing object, the printing setting conditions and the optimized raster image processor setting conditions as teacher data, and outputs the teacher data for machine learning of the learning completion model.
5. A printing data generating device, comprising: a machine learning unit having a first learning completion model and a second learning completion model, wherein the first learning completion model is a model obtained by performing machine learning using as input a relationship between data size of manuscript data of a plurality of printing jobs, printing setting conditions, and RIP setting conditions, and a printing speed when the manuscript data is printed by the printing unit, and the second learning completion model is a model obtained by performing machine learning using a set of data size of manuscript data of the arbitrary printing jobs, printing setting conditions, and optimized RIP setting conditions obtained by performing prediction processing and decision processing on a plurality of arbitrary printing jobs using the first learning completion model, and using as input a relationship between the data size and printing setting conditions of each set and the optimized RIP setting conditions of each set, wherein the prediction processing predicts the printing speed of the arbitrary printing job by inputting the data size of manuscript data of the arbitrary printing job, printing setting conditions, and arbitrary RIP setting conditions, and the decision processing determines the optimized RIP setting conditions so that the predicted printing speed approaches or exceeds the processing speed of the printing unit; an RIP setting condition determination unit that determines an optimized RIP setting condition for a print job of a print target using the second learned model; and a raster image processor processing unit that performs raster image processor processing of the print job of the print object based on the optimized raster image processor setting conditions determined by the raster image processor setting condition determination unit to generate print data, and outputs the print data to the printing unit, The printing speed refers to the number of printed sheets per unit time calculated based on the time from when a print job is received until when printed materials are discharged.
