Printed matter text proofreading method and device based on graphic recognition

Abstract

The invention relates to the technical field of text proofreading of printed matter, and specifically to a proofreading method and device for text in printed matter based on graphic recognition. The invention imparts metallic properties to text in printed matter by adding magnetic metal to special ink used for printing. Then, text images in printed matter are obtained through electrostatic imaging detection. The reliance on cumbersome manual or machine page turning operations in the traditional proofreading process is avoided, and the labor cost in the proofreading process is reduced. The areas corresponding to various printing defects of each text are obtained based on graphic recognition, and the accuracy of proofreading work is improved. The correlation of various printing defects is analyzed through correlation factors, and the possibility of co-occurrence of different defects under the same inducement is measured from a statistical perspective, reflecting the closeness between different printing defects due to the common inducement. According to the analysis results, potential problem sources are matched, production process optimization is performed, the quality of printed matter is improved, and positive feedback regulation is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of printed matter text proofreading, and in particular to a printed matter text proofreading method and device based on graphic recognition. Background Art

[0002] Proofreading printed materials is a key step to eliminate production errors and ensure the quality of printed materials. Through proofreading, we can ensure that the text, punctuation, pictures and other contents on the printed materials are completely consistent with the original manuscript, avoiding quality problems caused by errors in the typesetting and printing process.

[0003] Proofreading printed materials is directly related to the quality of the final product and the reader's reading experience. An effective proofreading process can reduce errors and improve the professionalism and accuracy of the text. With the development of science and technology, proofreading technology is also constantly improving. Modern proofreading technology not only relies on manual proofreading, but also combines a variety of technical means to improve the efficiency and accuracy of proofreading. For example: OCR text extraction technology, automatic typesetting system, etc.

[0004] However, existing printed text proofreading technology has significant deficiencies and limitations in many aspects. First, its recognition rate is still challenging, especially when dealing with subtle printing defects such as blurred text, stroke adhesion, broken strokes, and paper noise. The automatic detection capability of optical character recognition technology is insufficient, making it difficult to ensure high-precision proofreading results.

[0005] In addition, the traditional proofreading process often relies on cumbersome manual or machine page turning operations, requiring proofreaders to frequently check page by page. This process is time-consuming and labor-intensive, greatly reducing the overall efficiency of proofreading work. This repetitive mechanical labor not only increases labor costs, but also restricts the proofreading cycle of printed text.

[0006] Therefore, it is necessary to propose a proofreading method for printed text based on graphic recognition to solve the above problems. Summary of the invention

[0007] The purpose of the present invention is to solve the problems existing in the background technology and to propose a proofreading method for printed text based on graphic recognition;

[0008] The purpose of the present invention can be achieved through the following technical solutions:

[0009] In a first aspect, the present invention provides a method for proofreading printed text based on graphic recognition, which specifically comprises the following steps:

[0010] Step 1: preparing the sample to be tested;

[0011] The printing template of the printed product to be tested is accurately obtained and saved, and then a carefully formulated special magnetic ink is used to accurately print according to the pre-prepared printing template and bind into a book to prepare the printed product.

[0012] As a preferred method of the present invention, the printed matter is immersed in a solution rich in magnetic particles, followed by deep immersion and delicate shaking to ensure that each page of the printed matter is in full contact with the magnetic particles in the solution, thereby causing the magnetic particles to be uniformly adsorbed on the surface of the text in the printed matter, and the printed matter treated by the above process is recorded as a sample to be tested.

[0013] Step 2: Text perception and recording.

[0014] The sample to be tested is placed in a detection magnetic field with a preset strength. Based on electromagnetic induction, an electric current is generated in the magnetic particles in the sample to be tested, so that its surface is evenly covered with a layer of static charge.

[0015] Electrostatic imaging is performed on the sample to be tested layer by layer to generate electrostatic imaging of the magnetic particles in each layer, and the printed text images of each page of the sample to be tested are obtained and sorted by page number.

[0016] Step 3: Graphic recognition;

[0017] The printed text images of each page of the sample to be tested generated by electrostatic imaging are matched and overlapped with the image of the printing template according to their respective page numbers and recorded as overlapped images.

[0018] As a preferred embodiment of the present invention, seven types of printing defects are preset and numbered as follows: 1. disconnection defect; 2. short circuit defect; 3. rat bite defect; 4. fork defect; 5. shift defect; 6. splash defect; 7. pinhole defect.

[0019] It should be noted that in the category of printing defects, disconnection defects refer to the phenomenon that the parts of the text that should be printed continuously are broken, resulting in these parts not being closely connected as expected, destroying the integrity and readability of the text; short circuit defects are manifested as the accidental connection between two or more originally independent parts of the text, just like the short circuit in the circuit, which may be caused by factors such as excessive ink, layout design errors of the printing template or excessive printing pressure; mouse bite defects refer to the phenomenon of irregular missing along the edge of the inner or outer contour of the text, just like being gently gnawed by a small animal; fork defects refer to the accidental appearance of extra ink branches or protrusions on the outer contour edge of the text. Different from the short-circuit defect, the branch or protrusion does not connect two or more originally independent parts; the displacement defect is manifested as an overall displacement of part of the ink in the text relative to its expected position, which may be caused by the precision of the printing machinery, the change in the elasticity of the paper or the inaccurate positioning of the layout, resulting in uneven arrangement of the text; the spatter defect refers to the excess ink dots generated during the printing process, which may appear randomly around the text or on the layout; the pinhole defect refers to the small, round, irregular defects that do not intersect with the inner and outer contours that appear inside or around the text. They are like tiny holes and may be caused by material defects, uneven printing pressure or poor ink penetration.

[0020] Based on the graphic recognition, the areas S1ij, S2ij, S3ij, S4ij, S5ij, S6ij, and S7ij corresponding to the disconnection defect, short circuit defect, mouse bite defect, fork defect, displacement defect, splash defect, and pinhole defect of each word are obtained respectively in the overlapped pictures. Where i is the page positioning number, j is the word positioning number, i=1, 2, 3, ..., n; j=1, 2, ..., m. Where n is the total number of pages, and m is the total number of words in each page.

[0021] As a preferred embodiment of the present invention, an overall evaluation of printing quality is performed for the entire sample to be tested, specifically:

[0022] By formula Calculate the overall defect index E. Where X is the printing defect category number, X=1, 2, 3, 4, 5, 6, 7; where kX is the weight coefficient, kX=K1, K2, K3, K4, K5, K6 and k7; K1 to K7 are all natural numbers greater than 0, which are used to balance the weight of each coefficient in the formula calculation, thereby promoting the accuracy of the calculation result.

[0023] As a preferred embodiment of the present invention, for each character in the sample to be tested, a printing quality detail evaluation is performed, specifically:

[0024] By formula Calculate the character defect index eij for each character, where i is the page position number, j is the character position number, and X is the number of the printing defect category. Based on the value of the character defect index, the following color coding scheme is used to clearly distinguish its severity:

[0025] If the value of the character defect index eij of a text exists is greater than threshold 2, the text is marked in a striking red color to highlight that it has significant printing defects; if the value of the character defect index eij of a text exists is greater than threshold 1 and less than or equal to threshold 2, the text is marked in yellow, indicating that it has printing defects but has not yet reached the most serious level; if the value of the character defect index eij of a text exists is less than or equal to threshold 1, the text is marked in green to indicate that the text has relatively few printing defects or meets the printing standard. Threshold 2 and threshold 1 are both natural numbers greater than 0, which are used to divide the character defect index into three ranges according to the value size, and threshold 2 is greater than threshold 1.

[0026] As a preferred embodiment of the present invention, the printing defect correlation evaluation is performed based on the spatial distribution characteristics and statistical correlation of each printing defect category, specifically:

[0027] Extract the text marked in red and yellow and record it as defective text.

[0028] Get the original area Sij(0) of each defective text, get the area Sij(q) of various printing defects in each defective text, and use the formula

[0029] To calculate the probability of each specific defect occurring on the defect text , and the coexistence probability of each defect on the defect text .

[0030] The event X is defined as the occurrence of a specific defect, while event Y covers the occurrence of any other defects except X on the same defect text.

[0031] It is also worth pointing out that the original area Sij (0) specifically refers to the area of ​​the defective text that is uniquely corresponding to the page number positioning number i and the text positioning number j in the printing template, if it does not suffer from any printing defects and is presented after the standard printing process. It is the expected result obtained by assuming that the defective text is perfectly printed under ideal conditions. The areas Sij (X) of various printing defects are specifically Sij (1) to Sij (7), which are the areas of disconnection, short circuit, mouse bite, fork, shift, splash and pinhole defects in the defective text that is uniquely corresponding to the page number positioning number i and the text positioning number j.

[0032] As a preferred embodiment of the present invention, the concept of correlation factor is introduced. Represents the correlation between event X and event Y in the defect text uniquely corresponding to page location number i and text location number j. Event X is defined as the occurrence of a specific defect, while event Y covers the occurrence of any other defects except X on the same defect text.

[0033] By formula Calculate the correlation factor between event X and event Y in each defect text , , ..., .

[0034] Step 4: Analyze potential causes;

[0035] As a preferred method of the present invention, the correlation factor of each event in each defective text is obtained, the correlation factors for the same defective text are sorted and saved, and a correlation factor table is drawn.

[0036] The following relationship is assumed between the causes and printing defects:

[0037] Relationship 1: Excessive printing pressure leads to short circuit defects and shift defects;

[0038] Relationship 2: Local fracture of the printing template caused by mechanical stress or chemical corrosion leads to rat bite defects and disconnection defects;

[0039] Relationship 3: Poor contact leads to pinhole defects and rat bite defects;

[0040] Relationship 4: Excessive ink leads to splashing defects and fork defects;

[0041] Relationship 5: Electrostatic interference causes fork defects and pinhole defects.

[0042] Excessive printing pressure, local breakage of the printing template caused by mechanical stress or chemical corrosion, poor contact, excessive ink and electrostatic interference are named as causes A to E. Combined with the defect categories defined in step 3, they are numbered respectively, and the relationship is abbreviated as:

[0043] Relationship 1: Inducement A causes defect 2 + defect 5; Relationship 2: Inducement B causes defect 1 + defect 3; Relationship 3: Inducement C causes defect 3 + defect 7; Relationship 4: Inducement D causes defect 4 + defect 6; Relationship 5: Inducement E causes defect 4 + defect 7.

[0044] As a preferred method of the present invention, the cause tracing analysis is performed by comprehensively considering the correlation factors, specifically:

[0045] By formula

[0046] Calculate traceability characteristic parameters , , , and , where λ1 and λ2 are weight coefficients, which are used to balance the weights of various coefficients in the formula calculation, thereby improving the accuracy of the calculation results. Both λ1 and λ2 are natural numbers greater than 0.

[0047] When there is a traceability feature parameter greater than the preset threshold, the defective text that is considered to be uniquely corresponding to the page location number i and the text location number j is traced back to the corresponding cause.

[0048] Step 5: Adjustment and correction;

[0049] As a preferred embodiment of the present invention, when there is a traceability characteristic parameter greater than a preset threshold value, the printing template is obtained and the position of the defective text uniquely corresponding to the page number positioning number i and the text positioning number j is located, and the following targeted adjustments are performed at the position according to the inducement category to which the traceability characteristic parameter belongs:

[0050] When there are traceability characteristic parameters When it is greater than the preset threshold, the printing pressure at the position is reduced by , where α1 is the preset influence coefficient, β1 is the preset influence index, α1 and β1 are used together to adjust the influence of the traceability characteristic parameter in the formula calculation, and α1 and β1 are natural numbers greater than 0;

[0051] When there are traceability characteristic parameters When it is greater than the preset threshold, the position is detected by ultrasonic wave to obtain the fracture area of ​​the mold at that position. ,when When it is greater than the preset threshold, locate the traceability characteristic parameters greater than the preset threshold The corresponding page positioning number i and text positioning number j, and replace the corresponding font.

[0052] When there are traceability characteristic parameters When it is larger than a preset threshold value, cleaning treatment, corona treatment and flame treatment are performed at the position to increase its surface energy, thereby improving the adhesion of the ink at the position.

[0053] When there are traceability characteristic parameters When the value is greater than the preset threshold, the printing template is obtained and the position of the defective text that uniquely corresponds to the page number positioning number i and the text positioning number j is located, and the amount of ink applied is reduced at the position. , where α2 is the preset influence coefficient, β2 is the preset influence index, α2 and β2 are used together to adjust the influence of the traceability characteristic parameter in the formula calculation, and α2 and β2 are natural numbers greater than 0;

[0054] When there are traceability characteristic parameters When it is greater than a preset threshold, the printing template is obtained and the position of the defective text that uniquely corresponds to the page number positioning number i and the text positioning number j is located, and destaticization and cleaning treatment are performed at the said position.

[0055] The above targeted adjustments are traversed through all page number positioning numbers i and text positioning numbers j to complete the adjustment and correction of the printing process of the entire printing template

[0056] Step 6: Recheck the revised results;

[0057] As a preferred embodiment of the present invention, after completing step 5, the process automatically returns to step 1 and repeats step 3 in sequence. In this cycle, the overall defect index calculation method described in detail in step 3 is used to re-evaluate and calculate the corrected overall defect index. .

[0058] The corrected overall defect index Compare and analyze with the overall defect index E calculated before correction. If the overall defect index after correction shows a significant decrease, that is, its value is less than or equal to the E value before correction, the correction measures are considered effective, and the text proofreading work of the printed matter corresponding to the printing template has reached the expected standard, and it is determined to be completed. On the contrary, if the index rises instead of decreasing after correction, it is determined to be in an unfinished state.

[0059] In a second aspect, the present invention provides a proofreading device for printed matter text based on graphic recognition, which is used for deep immersion and fine shaking of printed matter, and comprises an equipment table, a first mechanical arm, a second mechanical arm, a solution tank and printed matter. The equipment table is equipped with a first mechanical arm, a second mechanical arm and a solution tank, wherein the solution tank is filled with the solution rich in magnetic particles.

[0060] The first robot arm and the second robot arm work together to accurately hold the printed matter and immerse it completely in the solution in the solution tank. Then, the two robot arms work together to achieve deep immersion of the printed matter in the solution, while also performing delicate shaking movements to ensure that the magnetic particles in the solution can fully contact the printed matter.

[0061] Compared with the prior art, the present invention has the following beneficial effects:

[0062] (1) The present invention imparts metallic properties to the text in the printed matter by adding magnetic metal to the special ink used for printing. At the same time, the text in the printed matter is slightly raised from the surface of the paper through the adsorption of magnetic particles in the solution, so that after the printed matter is hinged, a tiny air gap is naturally generated between each page, and the insulating properties of air are used to improve the precision and accuracy of subsequent electrostatic imaging detection. Then, the text image in the printed matter is obtained through electrostatic imaging detection. This avoids the reliance on cumbersome manual or machine page turning operations in the traditional proofreading process, reduces repetitive mechanical labor, and reduces the labor cost in the proofreading process;

[0063] (2) The present invention obtains the areas corresponding to the disconnection defect, short circuit defect, rat bite defect, fork defect, shift defect, splash defect and pinhole defect of each character based on graphic recognition, thereby improving the accuracy of proofreading work;

[0064] (3) The present invention analyzes the correlation of various printing defects through the correlation factor, measures the possibility of co-occurrence of different defects under the same inducement from a statistical perspective, and reflects the closeness between different printing defects due to the common inducement. According to the analysis results, the potential problem source is matched and the production process is optimized, which improves the quality of printed products and realizes positive feedback regulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings:

[0066] Figure 1 is a flow chart of the method of the present invention;

[0067] Figure 2 It is a schematic diagram of a printed text image of the present invention;

[0068] Figure 3 It is a schematic diagram of printing defects of the present invention;

[0069] Figure 4 It is a schematic diagram of the association factor table of the present invention;

[0070] Figure 5 It is a schematic diagram of the structure of the device for deep immersion and delicate shaking of the present invention.

[0071] Figure numerals: 1. Equipment table; 2. First robotic arm; 3. Second robotic arm; 4. Solution tank; 5. Printed material. DETAILED DESCRIPTION

[0072] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0073] See also Figure 1 As shown, the proofreading method of printed matter text based on graphic recognition specifically includes the following steps:

[0074] Step 1: preparing the sample to be tested;

[0075] The printing template of the printed product to be tested is accurately obtained and saved, and then a carefully formulated special ink is used to accurately print according to the pre-prepared printing template and bind it into a book to prepare the printed product.

[0076] It should be noted that the special ink contains metal particles of magnetic metals such as iron powder and nickel powder. By precisely controlling the ratio of these metal particles and the binder in the ink, it is ensured that the text printed using the special ink has the following characteristics: Feature 1: clear handwriting to meet reading requirements; Feature 2: unique magnetic properties to attract magnetic particles; Feature 3: stability and high strength to firmly adhere to the surface of the printed matter.

[0077] The printed matter is immersed in a solution rich in magnetic particles, followed by deep soaking and delicate shaking to ensure that each page of the printed matter is in full contact with the magnetic particles in the solution, thereby causing the magnetic particles to be uniformly adsorbed on the surface of the text in the printed matter. The printed matter treated by the above process is recorded as a sample to be tested.

[0078] This process aims to give the text in the printed matter a more significant metallic property, so that it can meet the needs of subsequent electrostatic imaging testing. At the same time, through the adsorption of magnetic particles, the text in the printed matter is slightly raised from the surface of the paper, so that after the printed matter is hinged, a small air gap is naturally generated between each page, and the insulating properties of air are used to improve the precision and accuracy of subsequent electrostatic imaging testing.

[0079] Step 2: Text perception and recording.

[0080] The sample to be tested is placed in a detection magnetic field with a preset strength. Based on electromagnetic induction, an electric current is generated in the magnetic particles in the sample to be tested, so that its surface is evenly covered with a layer of static charge.

[0081] Electrostatic imaging is performed on the sample to be tested layer by layer to generate electrostatic imaging of the magnetic particles in each layer.

[0082] See also Figure 2As shown, after electrostatic imaging, printed text images of each page of the sample to be tested are generated and sorted by page number.

[0083] Step 3: Graphic recognition;

[0084] The printed text images of each page of the sample to be tested generated by electrostatic imaging are matched and overlapped with the image of the printing template according to their respective page numbers and recorded as overlapped images.

[0085] There are seven preset categories of printing defects, which are numbered as follows: 1. disconnection defect; 2. short circuit defect; 3. rat bite defect; 4. fork defect; 5. shift defect; 6. spatter defect; 7. pinhole defect.

[0086] Please note that Figure 3 As shown in the figure, in the printing defect category, the disconnection defect refers to the breakage of the parts that should be printed continuously in the text, resulting in these parts not being closely connected as expected, destroying the integrity and readability of the text; the short circuit defect is manifested as an unexpected connection between two or more originally independent parts of the text, just like a short circuit in a circuit, which may be caused by factors such as excessive ink, layout design errors of the printing template, or excessive printing pressure; the mouse bite defect refers to the irregular missing phenomenon along the edge of the inner or outer contour of the text, just like being gently gnawed by a small animal; the fork defect refers to the unexpected appearance of extra ink branches or protrusions on the edge of the outer contour of the text. Different from short-circuit defects, the branches or protrusions do not connect two or more originally independent parts; displacement defects are manifested as an overall displacement of part of the ink in the text relative to its expected position, which may be caused by precision problems in the printing machinery, changes in the elasticity of the paper, or inaccurate positioning of the layout, resulting in uneven text arrangement; spatter defects refer to excess ink dots generated during the printing process, which may appear randomly around the text or on the layout; pinhole defects refer to small, round, irregular defects that do not intersect with the inner and outer contours that appear inside or around the text. They are like tiny holes and may be caused by material defects, uneven printing pressure, or poor ink penetration.

[0087] In the overlapped images, the areas S1ij, S2ij, S3ij, S4ij, S5ij, S6ij, S7ij, i.e., SXij, X=1, 2, 3, 4, 5, 6, 7, corresponding to the disconnection defect, short circuit defect, mouse bite defect, fork defect, displacement defect, splash defect and pinhole defect of each word are obtained based on graphic recognition, i.e., SXij, X=1, 2, 3, 4, 5, 6, 7. Where i is the page positioning number, j is the word positioning number, i=1, 2, 3, ..., n; j=1, 2, ..., m. Where n is the total number of pages, and m is the total number of words in each page.

[0088] For the entire sample to be tested, an overall evaluation of the printing quality is carried out, specifically:

[0089] By formula Calculate the overall defect index E. Where X is the printing defect category number, X=1, 2, 3, 4, 5, 6, 7; where kX is the weight coefficient, kX=K1, K2, K3, K4, K5, K6 and k7; K1 to K7 are all natural numbers greater than 0, which are used to balance the weight of each coefficient in the formula calculation, thereby promoting the accuracy of the calculation result.

[0090] For each word in the sample to be tested, the printing quality details are evaluated, specifically:

[0091] By formula Calculate the character defect index eij for each character, where i is the page position number, j is the character position number, and X is the number of the printing defect category. Based on the value of the character defect index, the following color coding scheme is used to clearly distinguish its severity:

[0092] If the value of the character defect index eij of a text exists is greater than threshold 2, the text is marked in a striking red color to highlight that it has significant printing defects; if the value of the character defect index eij of a text exists is greater than threshold 1 and less than or equal to threshold 2, the text is marked in yellow, indicating that it has printing defects but has not yet reached the most serious level; if the value of the character defect index eij of a text exists is less than or equal to threshold 1, the text is marked in green to indicate that the text has relatively few printing defects or meets the printing standard. Threshold 2 and threshold 1 are both natural numbers greater than 0, which are used to divide the character defect index into three ranges according to the value size, and threshold 2 is greater than threshold 1.

[0093] According to the spatial distribution characteristics and statistical correlation of each printing defect category, the printing defect correlation evaluation is carried out, specifically:

[0094] Extract the text marked in red and yellow and record it as defective text.

[0095] Get the original area Sij(0) of each defective text, get the area Sij(q) of various printing defects in each defective text, and use the formula

[0096] To calculate the probability of each specific defect occurring on the defect text , and the coexistence probability of each defect on the defect text .

[0097] The event X is defined as the occurrence of a specific defect, while event Y covers the occurrence of any other defects except X on the same defect text.

[0098] It is also worth pointing out that the original area Sij (0) specifically refers to the area of ​​the defective text that is uniquely corresponding to the page number positioning number i and the text positioning number j in the printing template, if it does not suffer from any printing defects and is presented after the standard printing process. It is the expected result obtained by assuming that the defective text is perfectly printed under ideal conditions. The areas Sij (X) of various printing defects are specifically Sij (1) to Sij (7), which are the areas of disconnection, short circuit, mouse bite, fork, shift, splash and pinhole defects in the defective text that is uniquely corresponding to the page number positioning number i and the text positioning number j.

[0099] In order to deeply understand the potential relationship between various defects, the concept of correlation factor is introduced. Represents the correlation between event X and event Y in the defect text uniquely corresponding to page location number i and text location number j. Event X is defined as the occurrence of a specific defect, while event Y covers the occurrence of any other defects except X on the same defect text.

[0100] By formula Calculate the correlation factor between event X and event Y in each defect text , , ..., .

[0101] It should be noted that in the printing process, the same cause often leads to multiple printing defects at the same time. For example, excessive printing pressure can cause short circuit defects and shift defects in the same defective text. The correlation factor focuses on measuring the possibility of co-occurrence of different defects under the same cause. The size of the correlation factor directly reflects the closeness between different printing defects due to the common cause. This analysis is of great significance for identifying potential problem sources in the printing process, optimizing production processes, and improving the quality of printed products.

[0102] Step 4: Analyze potential causes;

[0103] See also Figure 4 As shown, the correlation factor of each event in each defect text is obtained. The correlation factors for the same defect text are sorted and saved, and a correlation factor table is drawn.

[0104] The following relationship is assumed between the causes and printing defects:

[0105] Relationship 1: Excessive printing pressure leads to short circuit defects and shift defects;

[0106] Relationship 2: Local fracture of the printing template caused by mechanical stress or chemical corrosion leads to rat bite defects and disconnection defects;

[0107] Relationship 3: Poor contact leads to pinhole defects and rat bite defects;

[0108] Relationship 4: Excessive ink leads to splashing defects and fork defects;

[0109] Relationship 5: Electrostatic interference causes fork defects and pinhole defects.

[0110] Excessive printing pressure, local breakage of the printing template caused by mechanical stress or chemical corrosion, poor contact, excessive ink and electrostatic interference are named as causes A to E. Combined with the defect categories defined in step 3, they are numbered respectively, and the relationship is abbreviated as:

[0111] Relationship 1: Inducement A causes defect 2 + defect 5; Relationship 2: Inducement B causes defect 1 + defect 3; Relationship 3: Inducement C causes defect 3 + defect 7; Relationship 4: Inducement D causes defect 4 + defect 6; Relationship 5: Inducement E causes defect 4 + defect 7.

[0112] The cause tracing analysis is carried out by comprehensively considering the correlation factors, specifically:

[0113] By formula

[0114] Calculate traceability characteristic parameters , , , and , where λ1 and λ2 are weight coefficients, which are used to balance the weights of various coefficients in the formula calculation, thereby improving the accuracy of the calculation results. Both λ1 and λ2 are natural numbers greater than 0.

[0115] When there is a traceability feature parameter greater than the preset threshold, the defective text that is considered to be uniquely corresponding to the page location number i and the text location number j is traced back to the corresponding cause.

[0116] Step 5: Adjustment and correction;

[0117] When there is a traceability feature parameter greater than the preset threshold, the printing template is obtained and the position of the defective text that uniquely corresponds to the page number positioning number i and the text positioning number j is located, and the following targeted adjustments are performed at the position according to the cause category to which the traceability feature parameter belongs:

[0118] When there are traceability characteristic parameters When it is greater than the preset threshold, the printing pressure at the position is reduced by , where α1 is the preset influence coefficient, β1 is the preset influence index, α1 and β1 are used together to adjust the influence of the traceability characteristic parameter in the formula calculation, and α1 and β1 are natural numbers greater than 0;

[0119] When there are traceability characteristic parameters When it is greater than the preset threshold, the position is detected by ultrasonic wave to obtain the fracture area of ​​the mold at that position. ,when When it is greater than the preset threshold, locate the traceability characteristic parameters greater than the preset threshold The corresponding page positioning number i and text positioning number j, and replace the font of the corresponding text.

[0120] When there are traceability characteristic parameters When it is larger than a preset threshold value, cleaning treatment, corona treatment and flame treatment are performed at the position to increase its surface energy, thereby improving the adhesion of the ink at the position.

[0121] When there are traceability characteristic parameters When the value is greater than the preset threshold, the printing template is obtained and the position of the defective text that uniquely corresponds to the page number positioning number i and the text positioning number j is located, and the amount of ink applied is reduced at the position. , where α2 is the preset influence coefficient, β2 is the preset influence index, α2 and β2 are used together to adjust the influence of the traceability characteristic parameter in the formula calculation, and α2 and β2 are natural numbers greater than 0;

[0122] When there are traceability characteristic parameters When it is greater than a preset threshold, the printing template is obtained and the position of the defective text that uniquely corresponds to the page number positioning number i and the text positioning number j is located, and destaticization and cleaning treatment are performed at the said position.

[0123] The above targeted adjustments are traversed through all page number positioning numbers i and text positioning numbers j to complete the adjustment and correction of the printing process of the entire printing template

[0124] Step 6: Recheck the revised results;

[0125] After completing step 5, it automatically goes back to step 1 and repeats step 3. In this cycle, the overall defect index calculation method detailed in step 3 is used to re-evaluate and calculate the revised overall defect index. . This process is designed to comprehensively review and verify the effectiveness of corrective actions.

[0126] The corrected overall defect index Compare and analyze with the overall defect index E calculated before correction. If the overall defect index after correction shows a significant decrease, that is, its value is less than or equal to the E value before correction, the correction measures are considered effective, and the text proofreading work of the printed matter corresponding to the printing template has reached the expected standard, and it is determined to be completed. On the contrary, if the index rises instead of decreasing after correction, it is determined to be in an unfinished state.

[0127] See also Figure 5 As shown, a proofreading device for printed matter text based on graphic recognition is used for deep immersion and fine shaking of printed matter, and comprises an equipment table 1, a first robotic arm 2, a second robotic arm 3, a solution tank 4 and printed matter 5. The first robotic arm 2, the second robotic arm 3 and the solution tank 4 are installed on the equipment table 1, wherein the solution tank 4 is filled with the solution rich in magnetic particles.

[0128] The first robot arm 2 and the second robot arm 3 work together to accurately hold the printed matter 5 and work together to completely immerse it in the solution in the solution tank 4. Then, the printed matter 5 is deeply immersed in the solution through the precise coordination and operation of the two robot arms, and at the same time, a delicate shaking action is performed to ensure that the magnetic particles in the solution can fully contact the printed matter 5.

[0129] It should be understood that the terms "include" and "comprising" used in the specification and claims of the present disclosure indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0130] It should also be understood that the terms used in this disclosure are only for the purpose of describing specific embodiments and are not intended to limit the disclosure. As used in this disclosure and claims, the singular forms "a", "an", and "the" are intended to include plural forms unless the context clearly indicates otherwise. It should also be further understood that the term "and / or" used in this disclosure and claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations;

[0131] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for proofreading printed text based on graphic recognition, characterized in that: The following steps are involved: Step 1: preparing the sample to be tested; Obtain and save the printing template of the printed matter to be tested, then use special magnetic ink to accurately print and bind into a book to prepare the printed matter; immerse the printed matter in a solution rich in magnetic particles and perform deep immersion and delicate shaking to cause the magnetic particles to be uniformly adsorbed on the surface of the text in the printed matter, and record the printed matter processed by the above process as the sample to be tested; Step 2: text sensing and recording; placing the sample to be tested in a detection magnetic field with a preset strength, generating current in the magnetic particles in the sample to be tested based on electromagnetic induction, so that the surface of the sample is evenly covered with a layer of static charge; performing electrostatic imaging detection on the sample to be tested layer by layer, generating electrostatic imaging of the magnetic particles in each layer, obtaining the printed text images of each page of the sample to be tested and sorting them by page number; Step 3: Graphic recognition: The printed text images of each page of the sample to be tested generated by electrostatic imaging are matched and overlapped with the image of the printing template according to their respective page numbers, and recorded as overlapped images; A set of printing defect categories is preset; the areas corresponding to various printing defects of each character are obtained in the overlapped images based on graphic recognition; an overall printing quality assessment is performed on the entire sample to be tested, and an overall defect index is obtained by calculation; a detailed printing quality assessment is performed on each character in the sample to be tested, and a character defect index of each character is obtained by calculation: a printing defect correlation assessment is performed on the spatial distribution characteristics and statistical correlation of each printing defect category to obtain a set of correlation factors; Step 4: Potential cause analysis; obtain the correlation factor of each event in each defective text; set a preset relationship between the cause and the printing defect, and conduct cause tracing analysis by comprehensively considering the correlation factor: Step 5: Adjustment and correction: When there is a traceability characteristic parameter greater than the preset threshold value, obtain the printing template and locate the position of the defective text that uniquely corresponds to the page number positioning number and the text positioning number, and perform targeted adjustments at the position according to the cause category to which the traceability characteristic parameter belongs; Step 6, recheck the correction results; after completing step 5, automatically go back to step 1, and repeat step 3 in sequence; in this cycle, use the overall defect index calculation method detailed in step 3 to re-evaluate and calculate the corrected overall defect index; then compare and analyze the corrected overall defect index with the overall defect index calculated before correction; if the corrected overall defect index shows a significant decrease, that is, its value is less than or equal to the value before correction, it is considered that the correction measures are effective, and the text proofreading work of the printed material corresponding to the printing template has reached the expected standard, and it is determined to be completed; on the contrary, if the index rises instead of decreasing after correction, it is determined to be in an unfinished state.

2. The printed matter text proofreading method based on graphic recognition according to claim 1, characterized in that: A group of printing defects and their numbers are as follows: No. 1 is a disconnection defect, No. 2 is a short circuit defect, No. 3 is a rat bite defect, No. 4 is a fork defect, No. 5 is a shift defect, No. 6 is a spatter defect, and No. 7 is a pinhole defect.

3. The printed matter text proofreading method based on graphic recognition according to claim 2, characterized in that: The specific process of overall evaluation of printing quality is as follows: By formula Calculate the overall defect index E; Where X is the printing defect category number, X=1, 2, 3, 4, 5, 6, 7; where kX is the weight coefficient, kX=K1, K2, K3, K4, K5, K6 and k7; SXij is the area corresponding to the disconnection defect, short circuit defect, rat bite defect, fork defect, shift defect, splash defect and pinhole defect of each character, i is the page number positioning number, j is the character positioning number, i=1, 2, 3,..., n; j=1, 2,..., m, where n is the total number of pages, and m is the total number of characters on each page.

4. The method for proofreading printed matter text based on graphic recognition according to claim 3, characterized in that: The specific process of printing quality detail evaluation is as follows: By formula The character defect index eij of each word is calculated. According to the value of the character defect index, the following color coding scheme is used to clearly distinguish its severity: If the value of the character defect index eij of a text is greater than threshold value 2, the text will be marked in a striking red color to highlight the presence of significant printing defects; if the value of the character defect index eij of a text is greater than threshold value 1 and less than or equal to threshold value 2, the text will be marked in yellow, indicating that printing defects exist but have not yet reached the most serious level; if the value of the character defect index eij of a text is less than or equal to threshold value 1, the text will be marked in green to indicate that the text has relatively few printing defects or meets the printing standards; threshold value 2 and threshold value 1 are both natural numbers greater than 0, which are used to divide the character defect index into three ranges according to the numerical value, and threshold value 2 is greater than threshold value 1.

5. The method for proofreading printed matter text based on graphic recognition according to claim 4, characterized in that: The specific process of evaluating the correlation of printing defects is as follows: Extract the text marked in red and yellow and record it as defective text; Get the original area Sij(0) of each defective text, get the area Sij(q) of various printing defects in each defective text, and use the formula ; To calculate the probability of each specific defect occurring on the defect text , and the coexistence probability of each defect on the defect text ; The event X is defined as the occurrence of a specific defect, while event Y covers the occurrence of any other defect except X on the same defect text; The original area Sij(0) specifically refers to the area of ​​the defective text that is uniquely corresponding to the page number positioning number i and the text positioning number j in the printing template and is presented after the standard printing process if it does not suffer from any printing defects. It is the expected result obtained by assuming that the defective text is perfectly printed under ideal conditions; the areas of various printing defects Sij(X) are specifically Sij(1) to Sij(7), which are the areas of disconnection, short circuit, mouse bite, fork, shift, splash and pinhole defects in the defective text that is uniquely corresponding to the page number positioning number i and the text positioning number j respectively; Introducing the concept of relevance factor, relevance factor Represents the correlation between event X and event Y in the defect text uniquely corresponding to page location number i and text location number j; event X is defined as the occurrence of a specific defect, while event Y covers the occurrence of any other defects except X in the same defect text; By formula Calculate the correlation factor between event X and event Y in each defect text , , ..., .

6. The method for proofreading printed matter text based on graphic recognition according to claim 5, characterized in that: The specific analysis of preset relationships and cause tracing is as follows: Relationship 1: Excessive printing pressure leads to short circuit defects and shift defects; Relationship 2: Local fracture of the printing template caused by mechanical stress or chemical corrosion leads to rat bite defects and disconnection defects; Relationship 3: Poor contact leads to pinhole defects and rat bite defects; Relationship 4: Excessive ink leads to splashing defects and fork defects; Relationship 5: Electrostatic interference causes fork defects and pinhole defects; Excessive printing pressure, local rupture of the printing template caused by mechanical stress or chemical corrosion, poor contact, excessive ink and electrostatic interference are named as causes A to E respectively; they are numbered according to the defect categories defined in step 3, and the relationship is abbreviated as: Relationship 1: Inducement A leads to defect 2 + defect 5; Relationship 2: Inducement B leads to defect 1 + defect 3; Relationship 3: Inducement C leads to defect 3 + defect 7; Relationship 4: Inducement D leads to defect 4 + defect 6; Relationship 5: Inducement E leads to defect 4 + defect 7; By formula Calculate traceability characteristic parameters , , , and , where λ1 and λ2 are weight coefficients; When there is a traceability feature parameter greater than the preset threshold, the defective text that is considered to be uniquely corresponding to the page location number i and the text location number j is traced back to the corresponding cause.

7. The method for proofreading printed matter text based on graphic recognition according to claim 6, characterized in that: The targeted adjustments are specifically: When there are traceability characteristic parameters When it is greater than the preset threshold, the printing pressure at the position is reduced by , where α1 is the preset influence coefficient and β1 is the preset influence index; When there are traceability characteristic parameters When it is greater than the preset threshold, the position is detected by ultrasonic wave to obtain the fracture area of ​​the mold at that position. ,when When it is greater than the preset threshold, locate the traceability characteristic parameters greater than the preset threshold The corresponding page number positioning number i and text positioning number j, and replace the font of the corresponding text; When there are traceability characteristic parameters When it is greater than a preset threshold, cleaning treatment, corona treatment and flame treatment are performed at the location to increase its surface energy, thereby improving the adhesion of the ink at the location; When there are traceability characteristic parameters When the value is greater than the preset threshold, the printing template is obtained and the position of the defective text that uniquely corresponds to the page number positioning number i and the text positioning number j is located, and the amount of ink applied is reduced at the position. , where α2 is the preset influence coefficient and β2 is the preset influence index; When there are traceability characteristic parameters When it is greater than a preset threshold, the printing template is obtained and the position of the defective text that uniquely corresponds to the page number positioning number i and the text positioning number j is located, and a destaticizing and cleaning process is performed at the said position; The above-mentioned targeted adjustment is carried out through all page number positioning numbers i and text positioning numbers j to complete the adjustment and correction of the printing process of the entire printing template.

8. A printed matter text proofreading device based on graphic recognition, characterized in that: A method for proofreading printed text based on graphic recognition, applied to implement any one of claims 1 to 7, comprising: The immersion module immerses the printed matter in a solution rich in magnetic particles, followed by deep immersion and delicate shaking to ensure that each page of the printed matter is in full contact with the magnetic particles in the solution, thereby causing the magnetic particles to be uniformly adsorbed on the text surface of the printed matter.

9. The printed matter text proofreading device based on graphic recognition according to claim 8, characterized in that: Also includes: An equipment table (1), a first robotic arm (2), a second robotic arm (3), a solution tank (4) and a printed matter (5); the first robotic arm (2), the second robotic arm (3) and the solution tank (4) are installed on the equipment table (1), wherein the solution tank (4) is filled with the solution rich in magnetic particles; The first robotic arm (2) and the second robotic arm (3) work together to hold the printed matter (5) and to completely immerse it in the solution in the solution tank (4); then, through the cooperation and operation of the first robotic arm (2) and the second robotic arm (3), the printed matter (5) is deeply immersed in the solution, and a delicate shaking action is performed to ensure that the magnetic particles in the solution are in full contact with the printed matter (5).

Images