Method, equipment and storage medium for product packaging text review

Abstract

The present invention provides a commodity packaging text review method, which is used to compare whether the text content of a printed manuscript of a commodity packaging is consistent with the text content of an electronic manuscript of the commodity packaging. The commodity packaging text review method includes the following steps: S100) obtaining a picture of the printed manuscript and a picture of the electronic manuscript; S200) scanning the picture of the printed manuscript and the picture of the electronic manuscript by optical character recognition technology, and concentrating the text content in the picture of the printed manuscript and the picture of the electronic manuscript to obtain a printed manuscript text manuscript and an electronic manuscript text manuscript respectively; and S300) comparing the text on the printed manuscript text manuscript and the text on the electronic manuscript text manuscript to obtain a comparison result.

Description

Technical Field

[0001] The present invention relates to a commodity packaging review method, and in particular to a commodity packaging text review method, equipment and storage medium. Background Art

[0002] As the first step in the launch of new products, product packaging review, especially the text review on product packaging, is a very important step. Many consumer products often face the situation of a large number of new products being launched, and the peak sales season is fleeting. The overall speed of new products from creativity and design to production and market launch is very critical. Taking beverage packaging review as an example, when faced with tens of thousands of new product packaging material appearance changes, it is necessary to quickly replace and review the factory address, product information, supplier changes, etc. on labels, cartons, and film packages.

[0003] Traditional product packaging review is done through offline communication and manual visual review, which has the following pain points:

[0004] 1. A large amount of packaging material drawing review work requires the collaboration of several work departments, which results in high labor costs and low collaboration efficiency.

[0005] 2. Manual review is prone to errors and omissions. Manual review is tiring and boring, which leads to errors and omissions.

[0006] 3. The manuscript review time is long, which is not conducive to quickly responding to market business needs.

[0007] For packaging materials such as cartons and film bags with particularly large manuscript sizes and particularly small text areas, it is slow for people to locate text information and the operation is inconvenient. Summary of the invention

[0008] The embodiments of the present invention provide a method, device and storage medium for product packaging text review. The product packaging text review method uses artificial intelligence (AI) visual text recognition technology to assist in text review on product packaging, and quickly locates the text difference position and information of two manuscripts, thereby improving the quality and efficiency of product packaging text review work and effectively reducing the workload of manual identification of differences.

[0009] To achieve the above purpose, the first aspect of the embodiment of the present application proposes a method for reviewing the text of a product packaging, which is used to compare whether the text content of a printed manuscript of a product packaging is consistent with the text content of the electronic manuscript of the product packaging. The method for reviewing the text of a product packaging comprises the following steps:

[0010] S100) obtaining a picture of the printed manuscript and a picture of the electronic manuscript;

[0011] S200) scanning the image of the printed manuscript and the image of the electronic manuscript by optical character recognition technology, and collecting the text contents in the image of the printed manuscript and the image of the electronic manuscript to obtain a printed manuscript text manuscript and an electronic manuscript text manuscript respectively; and

[0012] S300) Compare the text on the printed manuscript and the text on the electronic manuscript to obtain a comparison result.

[0013] In some embodiments, step S200 includes the following steps:

[0014] S210) segmenting the image of the printed manuscript and the image of the electronic manuscript to obtain text areas of the image of the printed manuscript and the image of the electronic manuscript respectively; and

[0015] S240) respectively collect the text areas in the printed manuscript and the electronic manuscript to obtain the printed manuscript text draft and the electronic manuscript text draft respectively.

[0016] In some embodiments, step S200 further includes the following steps:

[0017] S230) marking the text areas in the printed manuscript and the electronic manuscript with text area marks;

[0018] Step S240 includes the following steps:

[0019] S241) identifying the text area mark in the image of the printed manuscript and the image of the electronic manuscript;

[0020] S242) respectively determining the spacing between adjacent text area marks in the image of the printed manuscript and the image of the electronic manuscript;

[0021] When step S242 determines that the distance between adjacent text area marks in the image of the printed manuscript or the image of the electronic manuscript is less than a preset distance, the following steps are performed:

[0022] S243) Merging the contents marked by adjacent text area marks whose spacing is less than the preset spacing in the image of the printed manuscript and the image of the electronic manuscript respectively, to obtain a printed manuscript text-concentrated image and an electronic manuscript text-concentrated image respectively.

[0023] In some embodiments, step S200 further includes the following steps:

[0024] S220) marking the non-text area in the printed manuscript and the electronic manuscript with a non-text area mark;

[0025] Step S240 further includes the following steps:

[0026] S244) cropping the area with the non-text area mark in the image in the printed manuscript text collection and the image in the electronic manuscript text collection to obtain the printed manuscript text manuscript and the electronic manuscript text manuscript.

[0027] In some embodiments, if step S242 determines that there are no adjacent text area marks with a spacing less than the preset spacing in the image of the printed manuscript or the image of the electronic manuscript, step S244 is directly performed.

[0028] In some embodiments, the product packaging text review method further comprises the following steps:

[0029] S400) splicing and displaying the comparison results for identification.

[0030] In some embodiments, step S300 includes the following steps:

[0031] S310) comparing line by line to match corresponding sentences in the printed manuscript text and the electronic manuscript text; and

[0032] S320) comparing corresponding sentences in the printed manuscript text and the electronic manuscript text;

[0033] If there are differences in the corresponding sentences compared in step S320, the following steps are performed:

[0034] S330) marking the difference marks at the difference locations;

[0035] If the corresponding sentences compared in step S320 are the same and there is no difference, the following steps are performed:

[0036] S340) Marking two corresponding sentences that do not have any differences as compared and without any differences.

[0037] To achieve the above-mentioned purpose, the second aspect of an embodiment of the present application proposes an electronic device, which includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method described in the first aspect is implemented.

[0038] To achieve the above-mentioned purpose, a third aspect of an embodiment of the present application proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in the first aspect is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0040] Figure 1 It is a flow chart of a commodity packaging text review method according to a preferred embodiment of the present invention.

[0041] Figure 2 The sub-steps of step S200 in the commodity packaging text review method according to the preferred embodiment of the present invention are explained.

[0042] Figure 3 The three-dimensional array converted from a picture during the commodity packaging text review method according to the preferred embodiment of the present invention is explained.

[0043] Figure 4 The sub-steps of sub-step S240 of step S200 in the commodity packaging text review method according to the preferred embodiment of the present invention are explained.

[0044] Figure 5 The manuscript marks formed in the commodity packaging text review method according to the preferred embodiment of the present invention are illustrated.

[0045] Figure 6 An example of the electronic manuscript text manuscript formed in the commodity packaging text review method according to the preferred embodiment of the present invention is shown.

[0046] Figure 7 An example of a physical manuscript text manuscript formed in the commodity packaging text review method according to the preferred embodiment of the present invention is provided.

[0047] Figure 8 The sub-steps of step S300 in the commodity packaging text review method according to the preferred embodiment of the present invention are explained.

[0048] Fig. 9 The splicing diagram formed in the commodity packaging text review method according to the preferred embodiment of the present invention is illustrated. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0050] It should be noted that, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0051] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0052] First, some nouns involved in this application are analyzed:

[0053] OCR (Optical Character Recognition) is a technology that converts text in paper documents into editable text.

[0054] Specification of the drawings Figure 1 A method for reviewing product packaging text according to a preferred embodiment of the present invention is explained. The method for reviewing product packaging text uses artificial intelligence (AI) visual text recognition technology to compare the text information in a printed manuscript of a product packaging and an electronic manuscript of the product packaging to determine whether the printed manuscript is correct, and includes the following steps:

[0055] S100) obtaining a picture of the printed manuscript and a picture of the electronic manuscript;

[0056] S200) scanning the image of the printed manuscript and the image of the electronic manuscript by optical character recognition technology, and collecting the text content in the image of the printed manuscript and the image of the electronic manuscript to obtain a printed manuscript text manuscript and an electronic manuscript text manuscript respectively;

[0057] S300) comparing the text on the printed manuscript and the text on the electronic manuscript to obtain a comparison result; and

[0058] S400) splices and displays the comparison results for easy identification.

[0059] According to the preferred embodiment of the present invention, step S200 scans the image of the printed manuscript and the image of the electronic manuscript line by line through optical character recognition technology, and excludes the non-text area of ​​the image of the printed manuscript and the image of the electronic manuscript. That is, the image of the printed manuscript and the image of the electronic manuscript are scanned through optical character recognition technology, and the non-text content of the image of the printed manuscript and the image of the electronic manuscript is excluded to obtain the text of the printed manuscript and the text of the electronic manuscript respectively.

[0060] According to the preferred embodiment of the present invention, the picture of the printed manuscript is obtained by taking a photo.

[0061] In addition to text content, most product packaging also includes non-text content such as graphics. For example, large-sized carton film packages have the characteristics of large size and small text, where the area occupied by the text content is small and scattered relative to the large size of the entire package. Step S200 excludes the non-text area of ​​the image of the printed manuscript and the image of the electronic manuscript through segmentation technology, and centrally arranges the scattered text areas for centralized comparison.

[0062] As shown in the attached figure of the specification Figure 2 As shown, step S200 includes the following steps:

[0063] S210) Segmenting the printed manuscript image and the electronic manuscript image;

[0064] S220) marking the non-text area in the printed manuscript and the electronic manuscript with a non-text area mark;

[0065] S230) marking the text area in the printed manuscript and the electronic manuscript with a text area mark;

[0066] S240) respectively concentrate the text areas in the printed manuscript and the electronic manuscript to obtain the printed manuscript text manuscript and the electronic manuscript text manuscript respectively.

[0067] In step S210, the printed manuscript image and the electronic manuscript image to be compared are read and read into the three-dimensional array structure of the following graphic expression using the opencv module. The graphics can be decomposed into a three-dimensional array vector, and each pixel value is in the range of 0 to 255. All subsequent algorithm processing is based on this array data. Its typical API interface includes:

[0068] cv2.imdecode(numpy.fromfile(img_path,dtype=numpy.uint8),-1) converts the image into a three-dimensional array. The graphic image is as follows Figure 3 shown.

[0069] Then, the printed three-dimensional array manuscript and the electronic three-dimensional array manuscript are scanned step by step by using OCR visual technology, wherein if a text area is recognized during the scanning process, it is retained, and if a non-text area is recognized, it is marked by leaving it black.

[0070] Then, the text areas retained in the printed manuscript and the electronic manuscript are respectively combined into one picture to form the printed manuscript text manuscript and the electronic manuscript text manuscript respectively.

[0071] Specifically, the printed manuscript and the electronic manuscript are segmented respectively. For packages with particularly large size and very small text area, it is inconvenient to review the text of the package due to the presence of a large amount of useless information, so it is necessary to segment the useful area of ​​the manuscript. The text area of ​​the manuscript itself is relatively small. The preferred embodiment of the present invention gradually scans the manuscript through OCR visual technology, retains the text area, deletes the text if it does not exist, and finally combines the recognized areas into a picture, which is the text area.

[0072] More specifically, the printed manuscript image and the electronic manuscript image are scanned step by step by OCR visual technology, wherein the text area is identified during the scanning process and marked. According to the preferred embodiment of the present invention, the text area is identified during the scanning process and marked with a red rectangular frame.

[0073] The text area in step S230 refers to an area containing text. According to the preferred embodiment of the present invention, step S220 marks the non-text area in the printed manuscript and the electronic manuscript by leaving black, wherein the area with text is marked as a concentrated text area with relatively concentrated text and a scattered text area with relatively scattered text, wherein the concentrated text area is presented as a bright area, wherein the scattered text area is marked by a preset marking symbol.

[0074] Step S240 merges the manuscript text regions by means of a frame merging method, wherein the marked text regions are merged.

[0075] According to the preferred embodiment of the present invention, step S240 merges the graphics with a large number of red rectangular boxes, wherein the small red rectangular boxes are merged, wherein a certain range is framed, and the spacing of the small rectangular boxes is judged, and the small rectangular boxes that are smaller than the preset range are merged into a large rectangular box. The process needs to determine which side to compare, and the position relationship of the two rectangular boxes can be determined by the coordinates of the upper left corner and the lower right corner, so as to determine their intersection, up and down, left and right orientation relationship. According to the preferred embodiment of the present invention, the preset range is 20 pixels.

[0076] Specifically, the text area in the image of the printed manuscript and the image of the electronic manuscript is circled and marked by a rectangular frame in step S230. Step S240 centrally processes the text area in the image of the printed manuscript and the image of the electronic manuscript according to the rectangular circle marks formed in step S230, so as to facilitate comparison in subsequent steps. Figure 4 , step S240 includes the following steps:

[0077] S241) identifying text area marks in the image of the printed manuscript and the image of the electronic manuscript; and

[0078] S242) Determine the spacing between adjacent text area marks in the image of the printed manuscript and the image of the electronic manuscript respectively.

[0079] When step S242 determines that the distance between adjacent text area marks in the image of the printed manuscript or the image of the electronic manuscript is less than a preset distance, the following steps are performed:

[0080] S243) merging the contents marked by adjacent text area marks whose spacing is less than the preset spacing in the image of the printed manuscript and the image of the electronic manuscript respectively, to obtain a printed manuscript text-concentrated image and an electronic manuscript text-concentrated image respectively; and

[0081] S244) cropping the area with the non-text area mark in the image in the printed manuscript text collection and the image in the electronic manuscript text collection to obtain the printed manuscript text manuscript and the electronic manuscript text manuscript.

[0082] Step S244 cuts out the area without text, and the printed manuscript text and the electronic manuscript text formed by cutting are both composed of recognizable text areas, and the text is concentrated, which is easy to be recognized and compared. In step S243, the contents of the adjacent text area marks with a spacing less than the preset spacing are merged, so that the text areas in the image of the printed manuscript and the image of the electronic manuscript are processed centrally, so as to facilitate the comparison in the subsequent steps.

[0083] According to the preferred embodiment of the present invention, step S242 determines the positional relationship between two adjacent text area marks based on the upper left corner and lower right corner coordinates of the adjacent text marks, and further determines their intersection, up, down, left, and right orientation relationships.

[0084] It is worth mentioning that if step S242 determines that there are no adjacent text area marks with a spacing less than the preset spacing in the image of the printed manuscript or the image of the electronic manuscript, step S244 is directly performed.

[0085] According to the preferred embodiment of the present invention, the text area is marked by a rectangular encircling mark, and the non-text area is marked by leaving black. In other words, the rectangular encircling mark is used as the text area mark, and the non-text area mark is left black.

[0086] In step S243, the contents circled by adjacent rectangular circle marks whose spacing is smaller than the preset spacing are merged, so that the text areas in the image of the printed manuscript and the image of the electronic manuscript are centrally processed, thereby facilitating comparison in subsequent steps.

[0087] According to this preferred embodiment of the present invention, step S242 determines the positional relationship between two adjacent rectangular encircled marks using the coordinates of the upper left corner and the lower right corner of the adjacent rectangular encircled marks, and further determines their intersection, up, down, left, and right orientation relationships.

[0088] It is worth mentioning that if step S242 determines that there are no adjacent rectangular circle marks with a spacing less than the preset spacing in the image of the printed manuscript or the image of the electronic manuscript, step S244 is directly performed.

[0089] Specification of the drawings Figure 5 The example of a rectangular encircling mark of an electronic manuscript packed in a carton is explained, wherein the text area of ​​the electronic manuscript is encircled by a rectangular frame. For easy identification, a rectangular frame with a preset color, such as red, can be used as the rectangular encircling mark.

[0090] The non-text areas in the printed manuscript and the electronic manuscript are marked by leaving them black.

[0091] Then, based on the recognition results, the bright area and the area with more text are cropped to obtain the recognizable text area, such as Figure 6 and Figure 7 shown.

[0092] Step S300 recognizes text by OCR technology and compares it line by line. According to the preferred embodiment of the present invention, the OCR text recognition technology in step S300 uses the paddleocr module to recognize clear text and outputs text line by line, based on which the text is recognized at close range.

[0093] Step S400 stitches the recognition and comparison results to form a stitching graph for display.

[0094] It is worth mentioning that during the text recognition and comparison process in step S300, the text information in the text area will be subjected to an intersection-type circular comparison. When the closest sentence is matched, it is determined that this sentence is the corresponding sentence. If there is a difference between the closest sentences, the difference is marked and displayed in the splicing diagram. If they are consistent, the two sentences are determined to be consistent.

[0095] Reference to the attached drawings of the specification Figure 8 , step S300 includes the following steps:

[0096] S310) comparing line by line to match corresponding sentences in the printed manuscript text and the electronic manuscript text; and

[0097] S320) comparing corresponding sentences in the printed manuscript text and the electronic manuscript text;

[0098] If there are differences in the corresponding sentences compared in step S320, the following steps are performed:

[0099] S330) marking the difference marks at the difference locations;

[0100] If the corresponding sentences compared in step S320 are the same and there is no difference, the following steps are performed:

[0101] S340) Marking two corresponding sentences that do not have any differences as compared and without any differences.

[0102] Specification of the drawings Fig. 9 An example of a mosaic graph outputted by step S400 is explained.

[0103] According to the preferred embodiment of the present invention, the spliced ​​image is composed of two manuscripts, wherein the upper manuscript of the spliced ​​image is the text area cut out from the electronic manuscript after OCR scanning and recognition, and the lower manuscript of the spliced ​​image is the text area cut out from the physical manuscript after being photographed and uploaded and then scanned and recognized by OCR. The two manuscripts are cross-compared line by line, and each line of the upper manuscript will match the most similar line of text in the lower manuscript as a comparison. When they are consistent, the line of text in the upper and lower manuscripts will be framed with a blue frame as a mark of no difference after comparison. If they are inconsistent, a red frame will be framed as a difference mark, and so on. The sub-table on the right side of the upper and lower manuscripts marks the recognized line text information in order and displays it line by line. In other words, step S300 also includes the following steps:

[0104] S350) Display the recognized text information line by line.

[0105] According to the preferred embodiment of the present invention, the image text area of ​​the printed manuscript can also be manually obtained by taking a photo or screenshot. The printed manuscript refers to a physical manuscript, that is, an actual product of the factory, which can be collected manually, that is, manually taking a photo of the text area of ​​the physical object and uploading it to the system for identification and comparison.

[0106] The preferred embodiment of the present invention utilizes OCR technology to scan large-size electronic manuscripts (film manuscripts) and physical manuscripts line by line, separate text areas from non-text areas, and finally cut out the text area at a specified position, and then perform text comparison to quickly locate the text difference position and information of the two manuscripts, effectively reducing the workload of manual identification of differences, liberating productivity, and reducing costs and increasing efficiency.

[0107] It is worth mentioning that the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including multiple instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store programs.

[0108] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the rights of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present invention should be within the scope of the rights of the present invention.

Claims

1. A method for reviewing the text of a product packaging, for comparing whether the text content of a printed manuscript of a product packaging is consistent with the text content of an electronic manuscript of the product packaging, characterized in that: The method comprises the following steps: S100) obtaining a picture of the printed manuscript and a picture of the electronic manuscript; S200) scanning the image of the printed manuscript and the image of the electronic manuscript by optical character recognition technology, and collecting the text contents in the image of the printed manuscript and the image of the electronic manuscript to obtain a printed manuscript text manuscript and an electronic manuscript text manuscript respectively; and S300) Compare the text on the printed manuscript and the text on the electronic manuscript to obtain a comparison result.

2. The method for reviewing product packaging text according to claim 1, characterized in that: Step S200 includes the following steps: S210) segmenting the image of the printed manuscript and the image of the electronic manuscript to obtain text areas of the image of the printed manuscript and the image of the electronic manuscript respectively; and S240) respectively collect the text areas in the printed manuscript and the electronic manuscript to obtain the printed manuscript text draft and the electronic manuscript text draft respectively.

3. The method for reviewing product packaging text according to claim 2, characterized in that: Step S200 further includes the following steps: S230) marking the text areas in the printed manuscript and the electronic manuscript with text area marks; Step S240 includes the following steps: S241) identifying the text area mark in the image of the printed manuscript and the image of the electronic manuscript; S242) respectively determining the spacing between adjacent text area marks in the image of the printed manuscript and the image of the electronic manuscript; When step S242 determines that the distance between adjacent text area marks in the image of the printed manuscript or the image of the electronic manuscript is less than a preset distance, the following steps are performed: S243) Merging the contents marked by adjacent text area marks whose spacing is less than the preset spacing in the image of the printed manuscript and the image of the electronic manuscript respectively, to obtain a printed manuscript text-concentrated image and an electronic manuscript text-concentrated image respectively.

4. The method for reviewing product packaging text according to claim 3, characterized in that: Step S200 further includes the following steps: S220) marking the non-text area in the printed manuscript and the electronic manuscript with a non-text area mark; Step S240 further includes the following steps: S244) cropping the area with the non-text area mark in the image in the printed manuscript text collection and the image in the electronic manuscript text collection to obtain the printed manuscript text manuscript and the electronic manuscript text manuscript.

5. The method for reviewing product packaging text according to claim 4, characterized in that: If step S242 determines that there are no adjacent text area marks with a spacing less than the preset spacing in the image of the printed manuscript or the image of the electronic manuscript, step S244 is directly performed.

6. The method for reviewing product packaging text according to any one of claims 1 to 5, characterized in that: Further comprising the steps of: S400) splicing and displaying the comparison results for identification.

7. The method for reviewing product packaging text according to any one of claims 1 to 5, characterized in that: Step S300 includes the following steps: S310) comparing line by line to match corresponding sentences in the printed manuscript text and the electronic manuscript text; and S320) comparing corresponding sentences in the printed manuscript text and the electronic manuscript text; If there are differences in the corresponding sentences compared in step S320, the following steps are performed: S330) marking the difference marks at the difference locations; If the corresponding sentences compared in step S320 are the same and there is no difference, the following steps are performed: S340) Marking two corresponding sentences that do not have any differences as compared and without any differences.

8. The method for reviewing product packaging text according to claim 7, characterized in that: Step S300 also includes the following steps: S350) Display the recognized text information line by line.

9. A computer device, characterized in that: The device includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory, used to store computer programs; The processor is used to implement the steps of the commodity packaging text review method described in any one of claims 1 to 8 when executing the program stored in the memory.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the commodity packaging text review method according to any one of claims 1 to 8 are implemented.

Images