WO2020147140A1 - 一种词码的生成方法、识别方法、装置、存储介质 - Google Patents

一种词码的生成方法、识别方法、装置、存储介质 Download PDF

Info

Publication number
WO2020147140A1
WO2020147140A1 PCT/CN2019/072818 CN2019072818W WO2020147140A1 WO 2020147140 A1 WO2020147140 A1 WO 2020147140A1 CN 2019072818 W CN2019072818 W CN 2019072818W WO 2020147140 A1 WO2020147140 A1 WO 2020147140A1
Authority
WO
WIPO (PCT)
Prior art keywords
word
target
code
sequence
word code
Prior art date
Application number
PCT/CN2019/072818
Other languages
English (en)
French (fr)
Inventor
李宝亮
Original Assignee
北京悦时网络科技发展有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京悦时网络科技发展有限公司 filed Critical 北京悦时网络科技发展有限公司
Priority to EP19910826.7A priority Critical patent/EP3913536A4/en
Priority to US17/413,008 priority patent/US11334780B2/en
Priority to JP2021541706A priority patent/JP7130881B2/ja
Publication of WO2020147140A1 publication Critical patent/WO2020147140A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/40—Processing or translation of natural language
    • G06F40/53—Processing of non-Latin text
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K19/00—Record carriers for use with machines and with at least a part designed to carry digital markings
    • G06K19/06—Record carriers for use with machines and with at least a part designed to carry digital markings characterised by the kind of the digital marking, e.g. shape, nature, code
    • G06K19/06009—Record carriers for use with machines and with at least a part designed to carry digital markings characterised by the kind of the digital marking, e.g. shape, nature, code with optically detectable marking
    • G06K19/06046—Constructional details
    • G06K19/06103—Constructional details the marking being embedded in a human recognizable image, e.g. a company logo with an embedded two-dimensional code
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/10—Text processing
    • G06F40/103—Formatting, i.e. changing of presentation of documents
    • G06F40/109—Font handling; Temporal or kinetic typography
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/22—Character recognition characterised by the type of writing
    • G06V30/224—Character recognition characterised by the type of writing of printed characters having additional code marks or containing code marks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/22—Character recognition characterised by the type of writing
    • G06V30/226—Character recognition characterised by the type of writing of cursive writing
    • G06V30/2264—Character recognition characterised by the type of writing of cursive writing using word shape
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/22—Character recognition characterised by the type of writing
    • G06V30/226—Character recognition characterised by the type of writing of cursive writing
    • G06V30/2268—Character recognition characterised by the type of writing of cursive writing using stroke segmentation

Definitions

  • the embodiment of the present invention relates to the field of machine vision recognition, in particular to a word code generation method, recognition method, device, and storage medium.
  • the two-dimensional code has the following problems in actual use: the two-dimensional code is machine language and can only be recognized by the machine. This means that using a QR code will take up space on the layout.
  • the form of the two-dimensional code is a dark square, which is visually obtrusive.
  • the embedded article and printed matter will obviously destroy the reading experience, which not only increases the difficulty of typesetting, but also requires high printing size and display accuracy.
  • the user does not know the intention of the QR code.
  • Two-dimensional codes have the above-mentioned shortcomings and are easily replaced by humans, which may cause various losses and problems.
  • the embodiments of the present invention provide a word code generation method, recognition method, device, and storage medium to solve the problems in the prior art.
  • the present invention provides a word code generation method, so that the same word sentence can generate a large number of word codes with the same characteristics of different machine vision.
  • the word codes are integrated with the text, and the characteristic is that the visual form is still text after generation, and machine recognition Later, the target file set by each can be called and the meaning of the word can be read by the human eye, and the human eye recognizes the machine recognition in one.
  • a number of split sequence elements are randomly selected for abnormal processing, but based on the abnormal sequence number and the target word sentence character sequence number, corresponding to the predetermined target file of the word code and generate words Based on the permutation and combination of different sequences, the realization of the same word sentence can generate a larger word code with different machine vision characteristics.
  • the visual form of the word code is still text. After scanning, different target files set by each can be called. Machine recognition and human Eye recognition unity.
  • randomly selecting several split sequence elements from the split sequence to perform abnormal processing to generate a word code including: sequentially recording the sequence numbers of the split sequence elements that are processed abnormally;
  • the word-sentence character sequence number and the sequence number of the different element generated according to the target word and sentence are combined to generate the word code serial number, and the word code serial number is respectively associated with the word code and the target file corresponding to the word code; for example, based on the code system Generating word codes directly generates readable word codes based on the rules designed by the system, including file types and response mechanisms, and the coded word codes are still implemented in each character split sequence.
  • performing different processing on the split sequence elements includes processing the attribute values of the split sequence elements, selecting sequences of different permutations and combinations and mixing dimensions, so that the same phrase can be generated Massive word codes with different machine vision features to correspond to express different target files.
  • the splitting of each text of the target phrase based on the connection points of the strokes includes:
  • the method further includes:
  • randomly selecting several split sequence elements from the split sequence to perform abnormal processing to generate a word code and further includes: sequentially recording the sequence numbers of the split sequence elements that are processed abnormally, and record them as the sequence numbers of the abnormal elements;
  • the present invention provides a word code generation device, including:
  • the target acquisition module is used to acquire the target words and sentences input by the user and the corresponding target file, and trigger the target splitting module;
  • the target splitting module is used to split each text of the target words and sentences based on the connection points of the strokes to obtain a split sequence and trigger the word code generation module;
  • the word code generation module is used to generate the word code based on the code system when the number of the split sequences is sufficient to express the characters of the target file, select part of the split sequence to add unusual features, based on the code system designed by the system Directly express the corresponding target file characters, file type and response mechanism; in the case that the number of split sequences is not enough to express the target file characters, randomly select several split sequence elements from the split sequence to perform abnormal processing, and generate The word code is to associate the word code with the target file based on the serial number rule of the system, and output the word code.
  • the present invention provides a word code recognition method, including:
  • the present invention provides a word code recognition device, including:
  • the target word and sentence recognition module is used to obtain an image containing the word code, identify the target word and sentence corresponding to the word code, and trigger the word code splitting module;
  • the word code splitting module is used to split each text in the target words and sentences according to the connection points of the strokes to obtain the word code splitting sequence and trigger the abnormal recognition module;
  • the abnormal recognition module is used to recognize the abnormal sequence from the word code splitting sequence and trigger the target file calling module;
  • the target file calling module is used to determine whether the word code is generated based on the code system. If it is, it will be directly read to call the target file; otherwise, the word will be called according to the target text serial number corresponding to the target word and sentence and the abnormal serial number.
  • the preset target file corresponding to the code is used to determine whether the word code is generated based on the code system. If it is, it will be directly read to call the target file; otherwise, the word will be called according to the target text serial number corresponding to the target word and sentence and the abnormal serial number.
  • the preset target file corresponding to the code is used to determine whether the word code is generated based on the code system. If it is, it will be directly read to call the target file; otherwise, the word will be called according to the target text serial number corresponding to the target word and sentence and the abnormal serial number.
  • the preset target file corresponding to the code is used to determine whether the word code is generated based on the code system. If it is, it will be directly read to call the target file; otherwise
  • the present invention provides a word code
  • the word codes include target words and sentences that have been processed differently
  • the target words and sentences of the abnormal processing are used to express the set target file
  • the target words and sentences are used to generate the word and sentence meaning serial numbers
  • the abnormally processed target words and sentences include abnormal elements
  • the unusual element is used to generate the serial number of the unusual element
  • the word and sentence character sequence number and the sequence number of the different element are used to generate the word code serial number to obtain a pre-saved target file corresponding to the word code serial number.
  • the present invention provides a computer-readable storage medium in which a program is stored, and the program is used to implement the above-mentioned word code generation method.
  • the present invention provides a computer-readable storage medium in which a program is stored, and the program is used to implement the word code recognition method as described above.
  • the visual form of the word code is text.
  • the same word sentence can generate a huge number of word codes with different machine vision characteristics. After scanning, it can read and call different target files set by each, and can express the meaning of the word. Machine recognition and human eye recognition are combined. One.
  • the shape is an abrupt dark square, and the embedded article and printed matter will have obvious foreign body feeling.
  • the word code can maintain smooth reading in various text scenes, and expand information for users. Connect to the service.
  • the combination of the word code and the text structure is easy to typeset, and it is difficult to be artificially replaced after application; the integration of human eye recognition and machine recognition means that there is no need to use scan codes to consume additional layout resources in order to achieve the scanning function.
  • scanning the six-word code area of "operation instruction video" in the printed materials of different manufacturers can open billions of video links to the operation instructions of specific products set by different manufacturers, which is extremely convenient for users For extended reading and connection, and the corresponding target file can be updated online from time to time; for example, by scanning different vocabulary of the same children’s book, you can open different links for knowledge query expansion, and so on.
  • the stepwise recognition can greatly avoid the occurrence of recognition errors in the first stage, and is not restricted by the complexity of the target file resource identifier .
  • FIG. 1 is a flowchart of a method for generating word codes according to an embodiment of the present invention
  • FIG. 2 is a structural diagram of a word code generation device provided by another embodiment of the present invention.
  • FIG. 3 is a flowchart of a word code recognition method provided by another embodiment of the present invention.
  • Fig. 4 is a structural diagram of a word code recognition device provided by another embodiment of the present invention.
  • 801 is a target word and sentence recognition module
  • 802 is a word code splitting module
  • 803 is an abnormal recognition module
  • 804 is a target file calling module.
  • the word code is integrated into the text structure, and the visual form is still text, which realizes the integration of machine recognition and human eye recognition without destroying the human eye recognition of the text.
  • a method for generating word codes is provided, as shown in FIG. 1, including:
  • Step 201 Obtain the target words and sentences input by the user and the corresponding target files;
  • the target words and sentences and target files input by the user are obtained, and the obtained target files are identified.
  • the target file can be a uniform resource identifier URL, or image information, audio information, text information,
  • Each target file type of video information and account information includes its response mechanism.
  • the system will automatically perform a validity check. If the pre-input target file can be opened after the check and judgment is read, the word code is output for the user to use, and the user can choose different sizes and formats; otherwise, the user is prompted to enter a valid The target file and regenerate the word code.
  • the format type of the target file is recognized, and the format type of the target file is determined to be readable, then save . Otherwise, the user is prompted to re-enter the file with a recognizable format type.
  • the target words and sentences and the corresponding target files when obtaining the target words and sentences and the corresponding target files, it also includes: generating a set of random numbers, generating and saving the meaning serial numbers of the words and sentences according to the random numbers in a preset manner, and specifically, the pronunciation or initials of each text corresponding to the target words and sentences can be pronounced
  • the abbreviation is concatenated with random numbers to generate a serial number for the meaning of words, sentences and characters.
  • the word and sentence meaning serial number is used to identify the current target word and sentence.
  • An example is taken to obtain the target words and sentences input by the user and the corresponding target files, and the target words and sentences are recognized, and the target words and sentences are recognized as the Chinese character "my phrase".
  • the target file is identified, and the target file is identified as a uniform resource identifier URL.
  • the acquired target words and sentences input by the user can be different settings of the target words and sentences by the user according to preset rules.
  • the system when the system obtains the target words and sentences, it only needs to judge it. If it is available, generate the word code; otherwise, prompt the user to set it as unavailable.
  • Step 202 Split each text of the target phrase based on the connection points of the strokes to obtain a split sequence
  • the method of splitting each text of the target word sentence based on the connection points of the strokes includes:
  • each text of the target word and sentence Recognize each text of the target word and sentence. If the recognized text can be split based on the stroke rules, the text will be split according to the stroke rules; otherwise, the text will be analyzed for connection points and split based on the connection points. Each element obtained after splitting is used as a split sequence. Further, according to the split sequence, the stroke sequence number of each element in the split sequence is established as the split sequence number.
  • the "my phrase” is divided into strokes, and a serial number is established for each stroke in sequence.
  • the " ⁇ " in the word The serial number of the stroke of " ⁇ ” is 1, the serial number of the stroke of " ⁇ ” in the word “ ⁇ ” is 2, the serial number of the stroke of " ⁇ ” in the word “ ⁇ ” is 3, and the stroke sequence of " ⁇ ” in the word “ ⁇ ”
  • the number is 4, the " ⁇ " stroke serial number in the word “I” is 5, the stroke serial number of " ⁇ ” in the word “ ⁇ ” is 6, and the stroke serial number of "Dian” in the word “ ⁇ ” is 7.
  • the stroke sequence number of " ⁇ ” in the character “ ⁇ ” is 8, and the stroke sequence number of " ⁇ ” in the character “ ⁇ ” is 9.
  • the rule of Hanzi stroke order, the sequence of each stroke in "My phrase” is performed Split and establish the corresponding stroke serial number.
  • the number of strokes of the four characters of "My Phrase" is 30, which can generate different word codes, which is far greater than the number of all mobile phone numbers in China.
  • a different combination of stroke sequences (such as single-scale bolding) is adopted for different combinations of stroke sequences.
  • n The stroke of n is n, it can produce Different kinds of "my phrases", that is, a huge number of A word code, All kinds of word codes can be corresponded and opened after recognition
  • target files such as links
  • the system will select different sequence numbers less than n to generate different word codes, mix different bolding ratios and different attribute values, so that the same word sentence can generate extremely large word codes with different machine vision characteristics, and scan the same Words and sentences can call their respective target files, such as scanning the same text "Zhangji Noodle House" of different merchants in different regions, you can get the connection, account information, content and various target files set by different merchants named Zhangji Noodle House .
  • the system when it is detected that the elements in the split sequence corresponding to the target word and sentence input by the user are less than the preset value, a variety of different mixtures can be used, the number of split sequences is increased, and the same split sequence is overlapped in different ways Method to expand the different word codes that can be generated. When the word codes corresponding to the special words are exhausted, the system will automatically prompt the user to increase the number of words to expand the number of combinations of word code serial numbers.
  • Step 203 Determine whether the number of split sequences is sufficient to express the characters of the target file, if yes, go to step 204; otherwise, go to step 205;
  • Step 204 Generate word codes based on the code system designed by this system, and select partial split sequences to add unusual features to directly express the corresponding target file characters;
  • the method of generating word codes is selected by judging whether the number of split sequences is sufficient to express the characters of the target document, and in the case that the number of split sequences is determined to be sufficient to express the characters of the target document, the word codes are generated based on the code system. In the case where it is determined that the number of split sequences is not enough to express the characters of the target file, the word code is generated based on the sequence number.
  • the word code generation mechanism selected by the user can also be obtained, and the word code is generated based on the word code generation mechanism selected by the user.
  • the user-selectable mechanism for generating word codes includes the mechanism for generating word codes based on the code system and the mechanism for generating word codes based on the serial number.
  • Step 205 randomly select a number of split sequence elements from the split sequence to perform different processing, generate a word code, associate the word code with the target file, and output the word code.
  • At least one element is randomly selected from the split sequence for different processing to obtain the processed element, and the processed element and other elements in the split sequence are recombined into the target according to the split sequence number.
  • the word sentence is obtained, and the word code is associated with the target file, so that when the word code is triggered, the target file associated with it can be obtained.
  • the method of the present invention When randomly selecting several split sequence elements from the split sequence for abnormal processing, it also includes obtaining all the abnormally processed elements, obtaining the split sequence number of each element, and adopting a preset method according to the split sequence number of each element Generate the sequence number of the different element, generate the serial number of the word code according to the serial number of the different element and the serial number of the word, sentence and character, associate the serial number of the word code with the word code and the target file, and save the word code serial number. Further, in order to ensure the uniqueness of the word code and the corresponding target file, the method of the present invention also includes automatically avoiding the existing word code serial number and processing method so as not to repeat.
  • the abnormal processing method includes, but is not limited to, adjusting at least one of the attribute values of thickness, stroke, shape, color value, and shape.
  • the attribute values corresponding to the different processing methods can also be set, so that the same different processing methods adopt different attribute values and the generated word code serial numbers are different. For example, based on the same target phrase, when the same element in the split sequence is bolded with different bolding ratios, the generated word code sequence numbers are different.
  • the number of word code sequence numbers has been increased, and the number of word codes can be generated based on the same target word sentence, which is the permutation and combination of all split sequences It can be mixed with different processing methods to greatly increase the number of combinations, that is, the same target phrase can generate the number of word codes with different characteristics.
  • the word code when the word code is associated with the target file, it also includes: scanning the word code to detect whether the set target file can be obtained, and if the target file can be obtained by verification, output the word code; otherwise, return to the current step Regenerate the word code.
  • the method of opening the target file corresponding to the phrase based on the sequence number of the word and sentence and the sequence number of the different element is based on the sequence number of the word and sentence and the sequence number of the different element.
  • stepwise recognition is that it can greatly avoid the occurrence of recognition errors in the first stage, and the background will automatically be automatically generated after the word code is generated. Scan to verify whether the set target file can be opened to ensure and improve the recognition accuracy.
  • a word code generation device is provided, as shown in FIG. 2, including:
  • the target obtaining module 401 is configured to obtain the target words and sentences input by the user and the corresponding target file, and trigger the target splitting module 402;
  • the target acquisition module 401 includes:
  • the target obtaining unit 4011 is used to obtain the target words and sentences input by the user and the corresponding target file, and trigger the file recognition unit 4012;
  • the file identification unit 4012 is used to identify the target file. In the case that the identified target file is a uniform resource identifier URL, trigger the URL verification unit 4013; when the identified target file is a non-uniform resource identifier URL In this case, trigger the readability check unit 4014;
  • the target file verification unit 4013 is used to verify the validity of the target file when the target file is identified, and trigger the file storage unit 4015 when it is determined that the target file is valid and can be opened; and when the target file is determined When invalid, prompt the user to enter a valid target file.
  • the readability verification unit 4014 is used to identify the format category of the target file when the identified target file is a non-uniform resource identifier URL, and verify the readability of the target file according to the format category, and When it is determined that the target file is readable, the file storage unit 4015 is triggered; when it is determined that the target file is not readable, the user is prompted to re-enter a file with a recognizable format type.
  • the file storage unit 4015 is used to store the target file.
  • the target acquisition module 401 further includes:
  • the character sequence code generation unit 4016 is connected to the file storage unit 4015, and is used to generate a set of random numbers, according to a preset method to generate and save the word and sentence meaning serial number according to the random number, and specifically can pronounce each character or initial letter corresponding to the target word and sentence
  • the abbreviation is concatenated with random numbers to generate a serial number for the meaning of words, sentences and characters.
  • the word and sentence meaning serial number is used to identify the current target word and sentence.
  • the target splitting module 402 is used to split each text of the target words and sentences based on the connection points of the strokes to obtain a split sequence, and trigger the word code generation module 403;
  • the target splitting module 402 includes:
  • the text recognition unit 4021 is used to recognize the target words and sentences and trigger the splitting unit 4022;
  • the splitting unit 4022 is used for splitting the text according to the stroke rules to obtain a split sequence when the recognized text can be split based on the stroke rules; it is also used for when the recognized text cannot be split based on the stroke rules. In the case of splitting, analyze the connection points of the text, split the connection points, and obtain the split sequence.
  • splitting unit 4022 specifically includes:
  • the stroke splitting sub-unit 40221 is used to split the target words and sentences according to the stroke rules when the recognized text can be split based on the stroke rules, and establish a split sequence number for each stroke in sequence, and split The serial number and the corresponding strokes are used as elements to form a split sequence.
  • connection point splitting subunit 40222 is used to analyze the connection points of the text when the recognized text cannot be split based on the stroke rules, split the connection points, and establish a split for each split part in sequence.
  • the sub-sequence number, the split sequence number and the corresponding split parts are used as elements to form the split sequence.
  • the word code generation module 403 is used to generate word codes based on the code system when the number of split sequences is sufficient to express the characters of the target file, and select part of the split sequences to add unusual features to directly express the corresponding target file characters; When the number of split sequences is not enough to express the characters of the target file, a number of split sequence elements are randomly selected from the split sequence to perform abnormal processing to generate a word code, associate the word code with the target file, and output the word code.
  • the word code generation module 403 includes:
  • the abnormality processing unit 4031 is configured to randomly select at least one element from the split sequence to perform abnormal processing to obtain the processed element, and recombine the processed element with other elements in the split sequence into a target phrase to obtain a word code.
  • the word code serial number generation unit 4032 is connected to the abnormality processing unit 4031, and is used to obtain all the abnormally processed elements when the number of split sequences is not enough to express the characters of the target file, and obtain the split sequence number of each element.
  • the preset method generates the sequence number of the different element according to the split sequence number of each element, and the sequence number of the word code according to the sequence number of the different element and the sequence number of the word meaning; it is also used to determine whether the word code serial number already exists, and the word code already exists
  • the abnormal processing unit 4031 is triggered to split the target sentence again; in the case that there is no word code serial number, the word code serial number is associated with the word code and the target file, and the word code sequence is saved number.
  • a code generation unit configured to generate word codes based on the code system when the number of split sequences is sufficient to express the characters of the target file, and select part of the split sequences to add unusual features to directly Express the corresponding target file characters;
  • the word code verification unit 4033 is respectively connected with the abnormality processing unit 4031 and the word code output unit 4034, and is used to scan the word code when the abnormality processing unit 4031 generates the word code to detect whether the target file can be obtained. If the target file is acquired within time, the word code output unit 4034 is triggered; otherwise, the abnormal processing unit 4031 is triggered to regenerate the word code;
  • the word code output unit 4034 is used to output word codes.
  • It also includes the addition of human eye identification tags when outputting the word code file to indicate that it can be scanned, limiting the output size and selecting specifications, and increasing the unusual feature of the small-size word code to maintain recognition.
  • the user can choose to re-output and select the style to customize .
  • a word code recognition method is provided, as shown in FIG. 3, including:
  • Step 601 Obtain an image containing the word code, and identify the target word and sentence corresponding to the word code;
  • the image containing the word code can be acquired by scanning, and the target word and sentence corresponding to the word code can be identified therefrom.
  • Step 602 Split each text in the target phrase according to the connection points of the strokes to obtain the word code split sequence
  • the characters of the target words and sentences are recognized.
  • the recognized characters can be split based on the stroke rules
  • the characters are divided into strokes according to the stroke rules, and each stroke is established in sequence.
  • Split the serial number use the split serial number and the corresponding strokes as elements to form a word code split sequence;
  • the text is analyzed for connection points to split the connection Point, and establish a split sequence number for each part obtained by splitting in order, and use the split sequence number and the corresponding split part as elements to form the word code splitting sequence.
  • Step 603 Identify the abnormal sequence from the word code splitting sequence
  • the standard text attributes of the target words and sentences in the current environment are obtained
  • the word code split sequence is judged according to the standard text attributes, and the split sequence elements that are different from the standard text attributes are filtered out, defined as abnormal elements, and their abnormal attributes are recorded in turn.
  • the abnormal recognition of the word code split sequence before performing the abnormal recognition of the word code split sequence, it also includes calculating the inclination angle of the word code based on the stroke projection in the application, and automatically compensates each stroke of the word code according to the calculation result to avoid scanning the word code due to graphics There is an inclination angle between the acquisition device and the word code, which causes the acquired word code image to be deformed (for example, when the word code is scanned forward, it will cause the scan to obtain the word code image to be thinner and thicker), resulting in abnormalities in the word code Errors or even failures occurred during recognition.
  • Step 604 Determine whether the word code is generated based on the code system, if yes, go to step 605; otherwise, go to step 606;
  • the abnormal sequence element at the pre-appointed position in the abnormal sequence is acquired, and it is determined whether the word code is generated based on the code system according to the abnormal sequence element at the pre-appointed position.
  • Step 605 Read directly and call the target file
  • the word code is generated based on the code system, so that there is no need to store the target file in a central server, saving resources At the same time, it is more convenient to read the word code and call the target file.
  • Step 606 According to the target text sequence and the abnormal sequence corresponding to the target words and sentences, call the preset target file corresponding to the word code.
  • the sequence number of the different element can be generated according to the split sequence number of the different element
  • the sequence number of the word code can be generated according to the serial number of the different element and the serial number of the word
  • sentence the target file corresponding to the word code serial number is obtained, and the target is called. file.
  • the word code serial number when generating the word code serial number, it also includes: judging whether the word code serial number exists, if yes, obtain the target file corresponding to the word code serial number, call the target file, and end; otherwise, prompt that the text is not a word code and end.
  • a word code recognition device is provided, as shown in FIG. 4, including:
  • the target word and sentence recognition module 801 is used to obtain an image containing the word code, identify the target word and sentence corresponding to the word code, and trigger the word code splitting module 802;
  • the target word and sentence recognition module 801 is used to obtain an image containing a word code by scanning, identify the target word and sentence corresponding to the word code therefrom, and trigger the word code splitting module 802.
  • the word code splitting module 802 is used to split each text in the target words and sentences according to the connection points of the strokes to obtain the word code splitting sequence, and trigger the abnormal recognition module 803;
  • the word code splitting module 802 is used to recognize each character of the target word and sentence.
  • the recognized character can be split based on the stroke rule
  • the word is split according to the stroke rule , And establish a split sequence number for each stroke in sequence, use the split sequence number and the corresponding stroke as elements to form the word code split sequence, and trigger the abnormal recognition module 803; the recognized text cannot be split based on the stroke rules
  • the abnormal recognition module 803 is used to recognize the abnormal sequence from the word code splitting sequence, and trigger the target file calling module 804;
  • the abnormal recognition module 803 is used to obtain the standard text attributes of the target words and sentences in the current environment, judge the word code splitting sequence according to the standard text attributes, and filter out elements with different attributes from the standard text. As a strange element, and record its sequence in turn.
  • the target file calling module 804 is used to determine whether the word code is generated based on the code system. If it is, it is directly read and the target file is called; otherwise, according to the target text sequence and the abnormal sequence corresponding to the target word and sentence, the preset corresponding to the word code is called target document.
  • the target file calling module 804 is used to directly read and call the target file when the word code is generated based on the code system designed by the system; used in the case of generating the word code based on the serial number , Generate the sequence number of the different element according to the split sequence number of the different element, generate the sequence number of the word code according to the serial number of the different element and the serial number of the word, sentence, character, obtain the target file corresponding to the word code serial number, and call the target file.
  • the target file calling module 804 is also used to determine whether the word code serial number exists when generating the word code serial number, and if yes, obtain the target file corresponding to the word code serial number, call the target file, and end; otherwise, prompt this text Non-word code, end.
  • a coding rule based on the system design that is, the code system, is provided to generate and read the word code.
  • the parallel direct system in the system can be selected. Generate word code mechanism to achieve, among them,
  • the method of generating word codes includes: splitting all the text of the target sentence based on the stroke nodes, and each splitting sequence is based on the specified abnormal characteristics, including bolding different proportions, shapes and various different attribute values, to sequentially correspond to commonly used characters.
  • the system will customize part of the split sequence to express the direct reading word code, and part of the sequence to express the target file type and response mechanism called after the word code is read.
  • the first stroke of the split sequence is bolded by 10%, representing letter a; the first stroke is bolded by 20%, representing letter b; the first stroke is bolded by 30%, representing letter c;
  • the code system is completed by analogy, so that each split sequence can be based on coding rules, using various attribute values of bold, stroke, thinning, shape, and various different ways to express and correspond to different characters, such as expression
  • the characters of is too long, the system can increase the split sequence, increase the abnormal points on each split sequence to accommodate more information, this mechanism maintains the font shape.
  • the method of identifying the word code includes: reading directly based on the code system or coding rule established by the system, the system will automatically activate the word code recognition mechanism when identifying the target word code, and then it can be read by identifying all the unusual features of the split sequence in turn Create a target file.
  • a word code in the fifth aspect of the present invention, includes target words and sentences that are processed differently;
  • the target phrase is used to generate the sequence number of the meaning of the phrase
  • the target words and sentences of the abnormal processing are used to express the set target file
  • the target words and sentences that have been processed differently include unusual elements
  • the strange element is used to generate the serial number of the strange element
  • the word-sentence-meaning serial number and the sequence number of the different element are used to generate the word code serial number to obtain the target file corresponding to the word code serial number.
  • the word code feature is realized based on the processing of the attribute value of the split sequence of each word and sentence, and it is integrated into the text structure, so that the same word sentence can generate a large number of word codes with different machine vision characteristics, which can be called after scanning.
  • the different target files set by each combine the expressions and meanings to provide users with information connection services;
  • the word code appears as a readable text, which means that no additional machine identification code is required to occupy the page resources, and the human eye recognition and machine recognition are integrated.
  • the word code only appears in the form of text.
  • the advantage is that it can be scanned and read by the machine to call the corresponding target file. It can also express the meaning of the text itself and can be smoothly embedded in articles, videos, pictures, and printed materials. , Outdoor scene.
  • a computer-readable storage medium is provided, and a program is stored in the computer-readable storage medium, and the program is used to implement the word code generation method described above.
  • a computer-readable storage medium is provided, and a program is stored in the computer-readable storage medium, and the program is used to implement the above-mentioned word code recognition method.

Abstract

一种词码的生成方法、识别方法、装置、存储介质，属于机器视觉识别领域。词码生成方法包括：输入目标词句及对应的目标文件，对目标词句的各文字基于笔画连接点进行拆分，得到拆分序列；随机选取若干拆分序列进行属性值异样处理，生成词码，与词码对应的目标文件建立关联，输出词码；词码识别方法：获取包含词码的图像，识别其中的目标词句；对目标词句按同一规则拆分后识别所有异样拆分序列，判断词码如基于系统设计的码制生成可直接读取调用，否则将根据词码对应的目标词句序列和异样序列，调用词码预先输入的目标文件，特点是同一词句可生成海量具备不同机器视觉特征的词码，词码形态仍为文字，人眼识别字义与机器识别合一。

Description

一种词码的生成方法、识别方法、装置、存储介质 技术领域
本发明实施例涉及机器视觉识别领域，具体涉及一种词码生成方法、识别方法、装置、存储介质。
背景技术
随着信息技术的发展，二维码、条形码以其编码范围广、信息容量大、使用方式简易等特点得到了广泛应用。
二维码在实际使用过程中存在以下问题：二维码是机器语言，仅能用机器识别。这意味着使用二维码会占用版面空间。二维码形态为深色方块，视觉突兀，嵌入文章及印刷物会明显破坏阅读体验，不仅增加了排版的难度，而且对印刷尺寸和显示精度要求较高。另外，使用时，如果不对其进行文字或其它形式的注释，用户不知二维码的意图。
二维码具有上述缺陷，且易被人为替换而造成各种损失及问题。
发明内容
为此，本发明实施例提供一种词码的生成方法、识别方法、装置、存储介质，以解决现有技术中存在的问题。
为了实现上述目的，本发明实施例提供如下技术方案：
第一方面，本发明提供一种词码的生成方法，让同一个词句可以生成海量具备不同机器视觉同特征的词码，词码融合于文字，特点是生成后视觉形态仍为文字，机器识别后可调用各自设定的目标文件又能人眼阅读其字义，人眼识别机器识别合一。
具体包括：
获取用户输入的目标词句及对应的目标文件；
对所述目标词句的各文字基于笔画连接点进行拆分，得到拆分序列；
判断拆分序列数量是否足以表达目标文件字符包括反应机制，是则基于系统设计的码制生成词码，选择部分拆分序列进行加粗、变细不同比例、笔触、形状各类属性值处理，基于不同异样特征来直接对应并表达目标文件的各字符、读取后的反应机制，文字拆分序列属性值调整又保持了词码的文字可读性；
否则从所述拆分序列中，随机选取若干拆分序列元素进行异样处理，但基于所述异样序列号及所述目标词句字义序列号，对应所述词码预先设定的目标文件并生成词码，基于不同序列的排列组合，实现同一个词句可以生成更庞大具备不同机器视觉特征的词码，词码视觉形态仍为文字，扫描后可调用各自设定的不同目标文件，机器识别与人眼识别合一。
在本发明另一实施例中，从所述拆分序列中随机选取若干拆分序列元素进行异样处理，生成词码，包括：依次记录被异样处理的拆分序列元素的序列号；
根据所述目标词句生成的词句字义序列号和异样元素序列号合并生成词码序列号，将所述词码序列号分别与所述词码及词码对应的目标文件建立关联；如基于码制生成词码，则直接基于系统设计的规则生成可读取词码，包括文件类型及反应机制，码制词码仍实现于各文字拆分序列。
在本发明另一实施例中，对所述拆分序列元素进行异样处理，包括对所述拆分序列元素的属性值进行处理，选择不同排列组合的序列及维度混合，让同一个词句可生成海量具备不同机器视觉特征的词码，以对应表达不同的目标文件。
本发明另一实施例中，所述的对目标词句的各文字基于笔画连接点进行拆分，包括：
对所述目标词句的各文字进行识别，在所述不同语言文字能够基于笔画规则拆分的情形下，根据笔画规则对所述文字进行拆分；在所述文 字不能够基于笔画规则拆分的情形下，对所述文字进行笔画连接点分析，基于连接点进行拆分，系统将保持同一拆分规则进行词码读取。
本发明另一实施例中，所述方法还包括，
根据所述目标词句生成词句字义序列号；
相应地，从所述拆分序列中随机选取若干拆分序列元素进行异样处理，生成词码，还包括：依次记录被异样处理的拆分序列元素的序列号，记做异样元素序列号；
根据所述词句字义序列号和所述异样元素序列号生成词码序列号，将所述词码序列号分别与所述词码及所述词码对应的目标文件建立关联；基于码制生成词码，则直接基于系统设计的码制生成可读取词码。
第二方面，本发明提供一种词码的生成装置，包括：
目标获取模块，用于获取用户输入的目标词句及对应的目标文件，并触发目标拆分模块；
目标拆分模块，用于对所述目标词句的各文字基于笔画连接点进行拆分，得到拆分序列，并触发词码生成模块；
词码生成模块，词码生成模块，用于在所述拆分序列数量足以表达目标文件字符的情形下，基于码制生成词码，选择部分拆分序列添加异样特征，基于系统设计的码制直接表达对应的目标文件字符、文件类型及反应机制；在所述拆分序列数量不足以表达目标文件字符的情形下，从所述拆分序列中随机选取若干拆分序列元素进行异样处理，生成词码，将所述词码与所述目标文件基于系统的序列号规则建立关联，输出词码。
第三方面，本发明提供一种词码的识别方法，包括：
获取包含词码的图像，识别所述词码对应的目标词句；
对所述目标词句中的各文字按笔画连接点进行拆分，得到词码拆分序列；
从所述词码拆分序列中识别出异样序列；
判断词码是否基于码制生成，是则直接读取，进行目标文件调用；否则根据所述目标词句对应的目标文字序列号和所述异样序列号，调用 所述词码对应的预先设定的目标文件。
第四方面，本发明提供一种词码的识别装置，包括：
目标词句识别模块，用于获取包含词码的图像，识别所述词码对应的目标词句，触发词码拆分模块；
词码拆分模块，用于对所述目标词句中各文字按笔画连接点进行拆分，得到词码拆分序列，触发异样识别模块；
异样识别模块，用于从所述词码拆分序列中识别出异样序列，触发目标文件调用模块；
目标文件调用模块，用于判断词码是否基于码制生成，是则直接读取，进行目标文件调用；否则根据所述目标词句对应的目标文字序列号和所述异样序列号，调用所述词码对应的预先设定的目标文件。
第五方面，本发明提供一种词码，
所述词码包括经异样处理的目标词句；
所述异样处理的目标词句用于表达设定的目标文件；
所述目标词句用于生成词句字义序列号；
所述经异样处理的目标词句中包括异样元素；
所述异样元素用于生成异样元素序列号；
所述词句字义序列号和所述异样元素序列号用于生成词码序列号，以获取与所述词码序列号对应的预先保存的目标文件。
第六方面，本发明提供一种计算机可读存储介质，所述计算机可读存储介质中存储有程序，所述程序用于实现如上所述的词码的生成方法。
第七方面，本发明提供一种计算机可读存储介质，所述计算机可读存储介质中存储有程序，所述程序用于实现如上所述的词码的识别方法。
本发明实施例具有如下优点：
词码视觉形态为文字，同一个词句可以生成数量庞大具备不同机器视觉特征的词码，即可扫描后读取调用各自设定的不同目标文件，又能表达字义，机器识别与人眼识别合一。
与二维码需要文字及各类注释，形态为突兀的深色方块，嵌入文章及印刷物会有明显异物感相比，词码能在各类文字场景保持顺畅阅读的同时，为用户拓展信息及连接服务。
另外，词码融合于文字结构易于排版，应用后被人为替换难度大；人眼识别与机器识别合一，意味着无需为了实现扫描功能，使用扫描码耗用额外版面资源。
在实际应用场景中，例如扫描不同厂商印刷品中“操作说明视频”这六个字的词码区域，可打开数十亿个不同厂商各自设定的具体产品的操作说明视频链接，用户可极便捷的进行扩展阅读及连接，且对应的目标文件可以时时在线更新；例如扫描同一儿童读物的不同词汇，可以打开不同链接进行知识查询扩展，以此类推。
在本发明中，基于词句字义序列号和异样元素序列号打开词组对应目标文件的方式，逐级识别可在第一阶段极大避免识别错误的发生，且不受目标文件资源标识符复杂度限制。
附图说明
图1为本发明一实施例提供的一种词码的生成方法流程图；
图2为本发明另一实施例提供的一种词码的生成装置结构图；
图3为本发明另一实施例提供的一种词码的识别方法流程图；
图4为本发明另一实施例提供的一种词码的识别装置结构图。
图中：801为目标词句识别模块、802为词码拆分模块、803为异样识别模块、804为目标文件调用模块。
具体实施方式
以下实施例用于说明本发明，但不用来限制本发明的范围。
词码融合于文字结构，视觉形态仍为文字，在不破坏文字人眼识别性的同时，实现机器识别与人眼识别合一。
在本发明的第一方面，提供一种词码的生成方法，如图1所示包括：
步骤201：获取用户输入的目标词句及对应目标文件；
在本发明实施例中，获取用户输入的目标词句和目标文件，对获取到的目标文件进行识别，其中，目标文件可以为统一资源标识符URL，还可以为图像信息、音频信息、文本信息、视频信息、账户信息各目标文件类型包括其反应机制。
系统会自动进行有效性校验，如果经过校验判定读取词码后，预先输入的目标文件能够打开，则输出词码供用户使用，用户可选不同尺寸和格式；否则提示用户输入有效的目标文件并重新生成词码。
在识别到目标文件为图像信息、音频信息、文本信息、视频信息、反应机制、账户信息中的一种的情形下，对目标文件进行格式类型识别，判定目标文件的格式类型可读，则保存。否则提示用户重新输入可识别格式类型的文件。
进一步，在获取到目标词句及对应目标文件时，还包括：生成一组随机数，按照预设方式根据随机数生成词句字义序列号并保存，具体可以将目标词句对应的各文字发音或首字母缩写于随机数拼接，生成词句字义序列号。词句字义序列号用于标识当前的目标词句。
进行举例说明，获取用户输入的目标词句和对应的目标文件，在对目标词句进行识别，识别到目标词句为汉字“我的词组”。对目标文件进行识别，识别到目标文件为统一资源标识符URL。
更进一步，获取到的用户输入的目标词句，可以是用户根据预设规则自行进行目标词句的异样设置，相对应的，系统在获取到该目标词句时，仅需对其进行判断，在判定其可用的情形下，生成词码；否则提示用户设置为不可用。
步骤202：对目标词句的各文字基于笔画连接点进行拆分，得到拆分序列；
在本发明实施例中，对目标词句的各文字基于笔画连接点进行拆分的方法，包括：
对目标词句的各文字进行识别，在识别到的文字能够基于笔画规则 拆分的情形下，则根据笔画规则对文字进行拆分；否则，对文字进行连接点分析，基于连接点进行拆分，将拆分后得到各个元素作为拆分序列。进一步地，根据拆分序列，建立拆分序列中各元素的笔画序列号，作为拆分序列号。
以目标词句为汉字中的“我的词组”进行举例说明：根据汉字笔画规则将“我的词组”进行笔画拆分，并且按照顺序依次为各笔画建立序列号，其中，“我”字中的“丿”笔画序列号为1，“我”字中的“一”笔画序列号为2，“我”字中的“亅”笔画序列号为3，“我”字中的“□”笔画序列号为4，“我”字中的“□”笔画序列号为5，“我”字中的“丿”笔画序列号为6，“我”字中的“丶”笔画序列号为7，“的”字中的“丿”笔画序列号为8，“的”字中的“丨”笔画序列号为9，按照此方法即汉子笔顺规则，依次为“我的词组”中的每个笔画进行拆分和建立对应的笔画序列号。
“我的词组”这四个字笔画数30，可产生不同词码数远大于中国所有手机号数量，对不同组合的笔画序列采取一种异样处理(譬如单一比例加粗)，其中进行异样处理的笔画为n，即可产生
Figure PCTCN2019072818-appb-000001
种不同的“我的词组”，即产生数量极其庞大的
Figure PCTCN2019072818-appb-000002
种词码，
Figure PCTCN2019072818-appb-000003
种词码均可对应并经识别后打开
Figure PCTCN2019072818-appb-000004
种链接等目标文件，系统会选择小于n不同序列数来生成不同词码，混合不同加粗比例及不同属性值，让同一词句能生成极其庞大的具备不同机器视觉特征的词码，实现扫描同一词句，可调用其各自设定的目标文件，例如扫描不同区域不同商户的同一文字“张记面馆”，可以获取不同名为张记面馆的商户设定的连接、账户信息、内容各类目标文件。
以目标词句为英文字母“My Phrase”进行举例说明：对英文字母进行连接点分析，基于规则拆分连接点，并且按照顺序依次建立拆分序列号，将拆分序列号及对应的拆分得到的部分作为元素，组成拆分序列，其它各国语言文字以此类推。
在本发明实施例中，检测到用户输入的目标词句对应的拆分序列中 的元素少于预设值时，可以采用多种异样混合、提高拆分序列数、同一拆分序列不同异样方式叠加的方法，以扩展可生成的不同词码。在特殊文字所对应的词码穷尽时，系统会自动提示用户增加文字数量以拓展词码序列号的组合数。
步骤203：判断拆分序列数量是否足以表达目标文件字符，是则执行步骤204；否则执行步骤205；
步骤204：基于本系统设计的码制生成词码，选择部分拆分序列添加异样特征，来直接表达对应的目标文件字符；
在本发明实施例中，通过判断拆分序列的数量是否足以表达目标文件字符来选择生成词码的方式，在判定拆分序列数量足以表达目标文件字符的情形下，基于码制生成词码。在判定拆分序列数量不足以表达目标文件字符的情形下，基于序列号生成词码。
本发明中，还可以获取用户选择的生成词码机制，基于该用户选择的生成词码机制生成词码。其中，用户可选择的生成词码机制包括，基于码制生成词码机制和基于序列号生成词码机制。在依据用户选择的基于码制生成词码的情形下，还需对拆分序列的数量进行判断，在拆分序列的数量足以表达目标文件字符的情形下，基于码制生成词码；否则提示用户基于序列号生成词码。
步骤205：从拆分序列中随机选取若干拆分序列元素进行异样处理，生成词码，将词码与目标文件建立关联，输出词码。
在本发明实施例中，从拆分序列中随机选取至少一个元素进行异样处理，得到处理后的元素，将处理后的元素与拆分序列中的其他元素，按照拆分序列号重新组合成目标词句，得到词码，将该词码与目标文件建立关联，以使得在词码被触发时，能够获取到与之建立关联的目标文件。
在从拆分序列中随机选取若干拆分序列元素进行异样处理时，还包括，获取所有经异样处理的元素，获取各元素的拆分序列号，采用预设方式根据各元素的拆分序列号生成异样元素序列号，根据异样元素序列 号和词句字义序列号生成词码序列号，将词码序列号与词码、目标文件建立关联，保存词码序列号。进一步，为了保证词码及对应的目标文件的唯一性，本发明方法还包括，自动避开已经存在词码序列号及处理方式以不重复。
在本发明实施例中，异样处理方式，包括但不限于调整粗细、笔触、形态、色值、形状中各属性值中的至少一种。更进一步，在进行异样处理时，还可以对异样处理方式对应的属性值进行设置，以使得同种异样处理方式采用不同的属性值，所生成的词码序列号不同。例如，基于同一目标词句，对拆分序列中相同的元素，采用不同的加粗比例进行加粗处理时，生成的词码序列号不同。
通过在元素粗细、色值、笔触、色值、形状各类属性值多维度方面的处理，增多了词码序列号的数量，基于同一目标词句可生成词码数量为所有拆分序列的排列组合数，可用不同异样处理方式混合，大幅提高组合数，即同一目标词句可生成具备不同特征的词码数量。
进一步，在词码与目标文件建立关联的情况下，还包括：扫描词码，检测是否能够获取到设定的目标文件，如果校验能获取到目标文件，则输出词码；否则返回当前步骤重新生成词码。
在本发明中，基于词句字义序列号和异样元素序列号打开词组对应目标文件的方式，逐级识别的优势在于可在第一阶段极大避免识别错误的发生，词码生成后后台会即刻自动扫描校验能否打开设置的目标文件，以保障和提高识别准确率。
在本发明的第二方面，提供一种词码的生成装置，如图2所示，包括：
目标获取模块401，用于获取用户输入的目标词句及对应目标文件，并触发目标拆分模块402；
在本发明实施例中，目标获取模块401，包括：
目标获取单元4011，用于获取用户输入的目标词句及对应目标文件，并触发文件识别单元4012；
文件识别单元4012，用于对目标文件进行识别，在识别到的目标文件为统一资源标识符URL的情形下，触发网址校验单元4013；在识别到的目标文件为非统一资源标识符URL的情形下，触发可读性校验单元4014；
目标文件校验单元4013，用于在识别到目标文件的情形下，对目标文件进行有效性校验，并在确定目标文件有效，能够被打开时，触发文件存储单元4015；并在确定目标文件无效时，提示用户输入有效的目标文件。
可读性校验单元4014，用于在识别到的目标文件为非统一资源标识符URL的情形下，对目标文件进行格式类别识别，根据格式类别对目标文件进行可读性校验，并在确定目标文件可读时，触发文件存储单元4015；在确定目标文件不可读时，提示用户重新输入可识别格式类型的文件。
文件存储单元4015，用于保存目标文件。
进一步地，目标获取模块401，还包括：
文字序列码生成单元4016，与文件存储单元4015连接，用于生成一组随机数，按照预设方式根据随机数生成词句字义序列号并保存，具体可以将目标词句对应的各文字发音或首字母缩写于随机数拼接，生成词句字义序列号。词句字义序列号用于标识当前的目标词句。
目标拆分模块402，用于对目标词句的各文字基于笔画连接点进行拆分，得到拆分序列，并触发词码生成模块403；
在本发明实施例中，目标拆分模块402，包括：
文字识别单元4021，用于对目标词句进行识别，并触发拆分单元4022；
拆分单元4022，用于在识别到的文字能够基于笔画规则拆分的情形下，根据笔画规则对文字进行拆分，得到拆分序列；还用于在识别到的文字不能够基于笔画规则拆分的情形下，对文字进行连接点分析，拆分连接点，得到拆分序列。
进一步地，拆分单元4022，具体包括：
笔画拆分子单元40221，用于在识别到的文字能够基于笔画规则拆分的情形下，根据笔画规则对目标词句进行笔画拆分，并且按照顺序依次为各笔画建立拆分序列号，将拆分序列号及对应的笔画作为元素，组成拆分序列。
连接点拆分子单元40222，用于在识别到的文字不能够基于笔画规则拆分的情形下，对文字进行连接点进行分析，拆分连接点，并按照顺序依次为各拆分得到部分建立拆分序列号，将拆分序列号及对应的拆分得到的部分作为元素，组成拆分序列。
词码生成模块403，用于在拆分序列数量足以表达目标文件字符的情形下，基于码制生成词码，选择部分拆分序列添加异样特征，来直接表达对应的目标文件字符；在所述拆分序列数量不足以表达目标文件字符的情形下，从拆分序列中随机选取若干拆分序列元素进行异样处理，生成词码，将词码与目标文件建立关联，输出词码。
在本发明实施例中，词码生成模块403，包括：
异样处理单元4031，用于从拆分序列中随机选取至少一个元素进行异样处理，得到处理后的元素，将处理后的元素与拆分序列中的其他元素重新组合成目标词句，得到词码。
词码序列号生成单元4032，与异样处理单元4031连接，用于在拆分序列数量不足以表达目标文件字符的情形下，获取所有经异样处理的元素，获取各元素的拆分序列号，采用预设方式根据各元素的拆分序列号生成异样元素序列号，根据异样元素序列号和词句字义序列号生成词码序列号；还用于判断是否已经存在词码序列号，在已经存在词码序列号的情形下，则触发异样处理单元4031，重新对目标词句进行拆分；在不存在词码序列号的情形下，将词码序列号与词码、目标文件建立关联，保存词码序列号。
本发明实施例中，还包括：码制生成单元，用于在所述拆分序列数量足以表达目标文件字符的情形下，基于码制生成词码，选择部分拆分 序列添加异样特征，来直接表达对应的目标文件字符；。
词码校验单元4033，分别与异样处理单元4031、词码输出单元4034连接，用于在异样处理单元4031生成词码的情形下，扫描词码，检测是否能够获取到目标文件，如果在预设时间内获取到目标文件，则触发词码输出单元4034；否则触发异样处理单元4031重新生成词码；
词码输出单元4034，用于输出词码。
还包括，输出词码文件时添加人眼识别标签以标明可扫描，限定输出尺寸及选择规格，对小尺寸词码进行异样特征增大以保持识别性，用户可选择重新输出、选择样式自定义。
在本发明的第三方面，提供一种词码的识别方法，如图3所示，包括：
步骤601：获取包含词码的图像，识别词码对应的目标词句；
在本发明实施例中，可以采用扫描的方式获取包含词码的图像，从中识别词码对应的目标词句。
步骤602：对目标词句中的各文字按笔画连接点进行拆分，得到词码拆分序列；
在本发明实施例中，对目标词句的各文字进行识别，在识别到的文字能够基于笔画规则拆分的情形下，则根据笔画规则对文字进行笔画拆分，并且按照顺序依次为各笔画建立拆分序列号，将拆分序列号及对应的笔画作为元素，组成词码拆分序列；在识别到的文字不能够基于笔画规则拆分的情形下，对文字进行连接点分析，拆分连接点，并按照顺序依次为拆分得到的各部分建立拆分序列号，将拆分序列号及对应的拆分得到的部分作为元素，组成词码拆分序列。
步骤603：从词码拆分序列中识别出异样序列；
在本发明实施例中，获取目标词句在当前环境中的标准文字属性；
根据标准文字属性对词码拆分序列进行判断，从中筛选出与标准文字属性不同的拆分序列元素，定义为异样元素，并依次记录其异样属性。
进一步，对词码拆分序列进行异样识别前还包括，在应用中对词码 基于笔画投影进行倾角计算，根据计算结果对词码的各笔画进行自动补偿，以避免扫描词码时，因图形采集设备与词码之间存在倾角，致使采集到的词码图像变形(例如，前倾扫描词码时，会造成扫描获取到的词码图像上细下粗)，造成在对词码进行异样识别时出现误差、甚至识别失败的情形。
步骤604：判断词码是否基于码制生成，是则执行步骤605；否则执行步骤606；
在本发明实施例中，获取异样序列中预先约定位置处的异样序列元素，根据该预先约定位置处的异样序列元素判断词码是否基于码制生成。
步骤605：直接读取，进行目标文件调用；
在本发明实施例中，在目标词句的拆分序列较多，足以为公众号或网址的目标文件的情形下，基于码制生成词码，使得无需把目标文件存在一个中央服务器里，节约资源，同时，读取词码调用目标文件更为便捷。
步骤606：根据目标词句对应的目标文字序列和异样序列，调用词码对应的预先设定的目标文件。
在本发明实施例中，可以根据异样元素的拆分序列号生成异样元素序列号，根据异样元素序列号和词句字义序列号生成词码序列号，获取词码序列号对应的目标文件，调用目标文件。
进一步，在生成词码序列号时，还包括：判断词码序列号是否存在，是则获取词码序列号对应的目标文件，调用目标文件，结束；否则，提示此文字非词码，结束。
在本发明的第四方面，提供一种词码的识别装置，如图4所示，包括：
目标词句识别模块801，用于获取包含词码的图像，识别词码对应的目标词句，触发词码拆分模块802；
在本发明实施例中，目标词句识别模块801，用于采用扫描的方式获取包含词码的图像，从中识别词码对应的目标词句，触发词码拆分模块 802。
词码拆分模块802，用于对目标词句中的各文字按笔画连接点进行拆分，得到词码拆分序列，触发异样识别模块803；
在本发明实施例中，词码拆分模块802，用于对目标词句的各文字进行识别，在识别到的文字能够基于笔画规则拆分的情形下，则根据笔画规则对文字进行笔画拆分，并且按照顺序依次为各笔画建立拆分序列号，将拆分序列号及对应的笔画作为元素，组成词码拆分序列，触发异样识别模块803；在识别到的文字不能够基于笔画规则拆分的情形下，对文字进行连接点分析，拆分连接点，并按照顺序依次为各拆分得到的各部分建立拆分序列号，将拆分序列号及对应的拆分得到的部分作为元素，组成词码拆分序列，触发异样识别模块803。
异样识别模块803，用于从词码拆分序列中识别出异样序列，触发目标文件调用模块804；
在本发明实施例中，异样识别模块803，用于获取目标词句在当前环境中的标准文字属性，根据标准文字属性对词码拆分序列进行判断，从中筛选出与标准文字属性不同的元素，作为异样元素，并依次记录其序列。
目标文件调用模块804，用于判断词码是否基于码制生成，是则直接读取，进行目标文件调用；否则根据目标词句对应的目标文字序列和异样序列，调用词码对应的预先设定的目标文件。
在本发明实施例中，目标文件调用模块804，用于在词码基于系统设计的码制生成的情形下，直接读取，进行目标文件调用；用于在基于序列号生成词码的情形下，根据异样元素的拆分序列号生成异样元素序列号，根据异样元素序列号和词句字义序列号生成词码序列号，获取词码序列号对应的目标文件，调用目标文件。
进一步，目标文件调用模块804，在生成词码序列号时，还用于判断词码序列号是否存在，是则获取词码序列号对应的目标文件，调用目标文件，结束；否则，提示此文字非词码，结束。
在本发明实施例中，提供一种基于系统设计的编码规则，即码制来生成和读取词码的方式，针对词码拆分序列足以表达目标文件的情况，可选择系统中并行的直接生成词码机制来实现，其中，
生成词码方法包括：将目标词句所有文字基于笔画节点进行拆分，每一个拆分序列都基于规定的异样特征，包括加粗不同比例、形状及各类不同属性值，来依次对应常用字符，从而表达设置的目标文件，系统会自定义部分拆分序列来表达直读词码、部分序列来表达该词码读取后调用的目标文件类型、反应机制。例如：文字基于笔画节点拆分后，拆分序列第1笔加粗10％，代表字母a；第1笔加粗20％，代表字母b；第1笔加粗30％，代表字母c；以此类推完成码制编排，实现每1个拆分序列都可基于编码规则，使用加粗、笔触、变细、形状各类属性值，各种异样方式相互组合来表达和对应不同字符，如表达的字符过长，系统可增加拆分序列、增加各拆分序列上的异样点以容纳更多信息，此种机制保持了字形。
识别词码的方法包括：基于系统制定的码制即编码规则来直接读取，系统在识别目标词码时会自动启用该词码识别机制，依次识别所有拆分序列的异样特征就可读取出一个目标文件。
在本发明第五方面，提供一种词码，词码包括经异样处理的目标词句；
目标词句用于生成词句字义序列号；
所述异样处理的目标词句用于表达设定的目标文件；
经异样处理的目标词句中包括异样元素；
异样元素用于生成异样元素序列号；
词句字义序列号和异样元素序列号用于生成词码序列号，以获取与词码序列号对应的目标文件。
在本发明实施例中，词码特点是基于词句各文字拆分序列属性值处理来实现，融合于文字结构，让同一个词句可以生成海量具备不同机器 视觉特征的词码，实现扫描后可调用各自设定的不同目标文件，又把表达字义融为一体，为用户提供信息连接服务；
不破坏文字人眼识别性，词码作为仍可阅读的文字出现，意味着不再需要额外的机器识别码占用版面资源，人眼识别机器识别合一。
与二维码等方式不同，词码仅以文字形态出现，优点为即可被机器扫描读取调用对应的目标文件，又能表达文字本身的字义，可顺畅的嵌入文章、视频、图片、印刷品、户外场景。
在本发明的第六方面，提供一种计算机可读存储介质，计算机可读存储介质中存储有程序，程序用于实现如上所述的词码的生成方法。
在本发明的第七方面，提供一种计算机可读存储介质，计算机可读存储介质中存储有程序，程序用于实现如上所述的词码的识别方法。
虽然，上文中已经用一般性说明及具体实施例对本发明作了详尽的描述，但在本发明基础上，可以对之作一些修改或改进，这对本领域技术人员而言是显而易见的。因此，在不偏离本发明精神的基础上所做的这些修改或改进，均属于本发明要求保护的范围。
虽然，上文中已经用一般性说明及具体实施例对本发明作了详尽的描述，但在本发明基础上，可以对之做一些修改或改进，这对本领域技术人员而言是显而易见的。因此，在不偏离本发明精神的基础上所做的这些修改或改进，均属于本发明要求保护的范围。

Claims (9)

  1. 一种词码的生成方法，其特征在于，包括：
    获取用户输入的目标词句及对应的目标文件；
    对所述目标词句的各文字基于笔画连接点进行拆分，得到拆分序列；
    判断拆分序列数量是否足以表达目标文件字符及反应机制，是则基于本系统设计的码制生成词码，选择部分拆分序列进行加粗、变细不同比例各类属性值异样组合编码处理，基于不同机器视觉异样特征，来直接对应并表达目标文件的字符及反应机制；
    否则从所述拆分序列中，随机选取若干拆分序列元素进行属性值异样处理，但基于所述异样序列及所述目标词句字义序列，对应所述词码预先设定的目标文件并生成词码，基于不同拆分序列的排列组合，实现同一个词句可以生成海量具备不同机器视觉特征的词码；词码特点是融合于文字，文字拆分序列属性值异样调整保持了词码的文字可读性，视觉形态仍为文字，既能扫描后基于异样特征调用各自设定的不同目标文件，又能表达字义，机器识别与人眼识别合一。
  2. 如权利要求1所述的方法，其特征在于，从所述拆分序列中随机选取若干拆分序列元素进行属性值异样处理，生成词码，包括：依次记录被异样处理的拆分序列元素的序列号包括反应机制；
    根据所述目标词句生成的词句字义序列号和异样元素序列号合并生成词码序列号，将所述词码序列号分别与所述词码及词码对应的目标文件建立关联；如基于码制生成词码，则直接基于系统设计的规则生成可读取词码，包括反应机制，码制词码仍实现于文字构架。
  3. 如权利要求1所述的方法，其特征在于，对所述拆分序列元素进行的属性值进行处理，选择不同排列组合的序列及属性值维度进行组合，会让同一个词句生成数量更庞大的具备不同机器视觉意义的词码，以对应表达不同的目标文件。
  4. 一种词码的生成装置，其特征在于，包括：
    目标获取模块，用于获取用户输入的目标词句及对应的目标文件，并触发目标拆分模块；
    目标拆分模块，用于对所述目标词句的各文字基于笔画连接点进行拆分，得到拆分序列，并触发词码生成模块；
    词码生成模块，用于在所述拆分序列数量足以表达目标文件字符的情形下，基于码制生成词码，选择部分拆分序列添加异样特征，基于系统设计的码制直接表达对应的目标文件字符、反应机制；在所述拆分序列数量不足以表达目标文件字符的情形下，从所述拆分序列中随机选取若干拆分序列元素进行异样处理，生成词码，将所述词码与所述目标文件基于目标文字序列号及异样序列号建立关联，输出词码。
  5. 一种词码的识别方法，其特征在于，包括：
    获取包含词码的图像，识别所述词码对应的目标词句；
    对所述目标词句中的各文字按笔画连接点进行拆分，得到词码拆分序列；
    从所述词码拆分序列中识别出异样序列；
    判断词码是否基于系统设计的码制生成，是则直接读取，进行目标文件调用；否则根据所述目标词句对应的目标文字序列和所述异样序列，调用所述词码对应的预先设定的目标文件。
  6. 一种词码的识别装置，其特征在于，包括：
    目标词句识别模块，用于获取包含词码的图像，识别所述词码对应的目标词句，触发词码拆分模块；
    词码拆分模块，用于对所述目标词句中各文字按笔画连接点进行拆分，得到词码拆分序列，触发异样识别模块；
    异样识别模块，用于从所述词码拆分序列中识别出异样序列，触发目标文件调用模块；
    目标文件调用模块，用于判断词码是否基于系统设计的码制生成，是则直接读取，基于反应机制进行目标文件调用；否则根据所述目标词句对应的目标文字序列和所述异样序列，调用所述词码对应的预先设定 的目标文件。
  7. 一种词码，其特征在于，
    所述词码包括经异样处理的目标词句；
    所述异样处理的目标词句用于表达设定的目标文件；
    所述目标词句用于生成词句字义序列号；
    所述经异样处理的目标词句中包括异样元素；
    所述异样元素用于生成异样元素序列号；
    所述词句字义序列号和所述异样元素序列号用于生成词码序列号，以获取与所述词码序列号对应的预先保存的目标文件。
  8. 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质中存储有程序，所述程序用于实现如权利要求1-3所述的词码的生成方法。
  9. 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质中存储有程序，所述程序用于实现如权利要求5所述的词码的识别方法。
PCT/CN2019/072818 2019-01-17 2019-01-23 一种词码的生成方法、识别方法、装置、存储介质 WO2020147140A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
EP19910826.7A EP3913536A4 (en) 2019-01-17 2019-01-23 PHRASE CODE GENERATING METHOD AND APPARATUS, PHRASE CODE VERIFICATION METHOD AND APPARATUS, AND RECORDING MEDIA
US17/413,008 US11334780B2 (en) 2019-01-17 2019-01-23 Method for generating word code, method and device for recognizing codes
JP2021541706A JP7130881B2 (ja) 2019-01-17 2019-01-23 ワードコードを生成する方法、ワードコードを認識する方法、及びその装置、コンピュータープログラム

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910045595.5 2019-01-17
CN201910045595.5A CN109766978B (zh) 2019-01-17 2019-01-17 一种词码的生成方法、识别方法、装置、存储介质

Publications (1)

Publication Number Publication Date
WO2020147140A1 true WO2020147140A1 (zh) 2020-07-23

Family

ID=66452466

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/072818 WO2020147140A1 (zh) 2019-01-17 2019-01-23 一种词码的生成方法、识别方法、装置、存储介质

Country Status (5)

Country Link
US (1) US11334780B2 (zh)
EP (1) EP3913536A4 (zh)
JP (1) JP7130881B2 (zh)
CN (1) CN109766978B (zh)
WO (1) WO2020147140A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113343639A (zh) * 2021-05-19 2021-09-03 网易（杭州）网络有限公司 产品标识码图生成、基于产品标识码图的信息查询方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101763516A (zh) * 2010-01-15 2010-06-30 南京航空航天大学 一种基于拟合函数的文字识别方法
CN101908290A (zh) * 2010-07-27 2010-12-08 李水超 英语读物及其识读机
CN102902968A (zh) * 2011-07-25 2013-01-30 上海博路信息技术有限公司 一种手机扫描快速获取出版物内容的方法
US20170090693A1 (en) * 2015-09-25 2017-03-30 Lg Electronics Inc. Mobile terminal and method of controlling the same
CN108830126A (zh) * 2018-06-20 2018-11-16 上海凌脉网络科技股份有限公司 一种基于图像智能识别的产品营销互动方法

Family Cites Families (28)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3832686A (en) * 1971-02-25 1974-08-27 I Bilgutay Bar code font
US3990043A (en) * 1971-12-30 1976-11-02 Xerox Corporation Character coding and recognition system
JPH07192091A (ja) * 1993-12-27 1995-07-28 Oki Electric Ind Co Ltd オンライン手書き文字列切出し装置
JP2836579B2 (ja) * 1996-04-19 1998-12-14 日本電気株式会社 文字切り出し候補生成装置
WO2002015004A2 (en) * 2000-08-14 2002-02-21 Transvirtual Technologies, Inc. Portable operating environment for information devices
JP2003259112A (ja) * 2001-12-25 2003-09-12 Canon Inc 透かし情報抽出装置及びその制御方法
CN1499357A (zh) * 2002-11-01 2004-05-26 ���Ծ 字词联体标注方法及其字模与字图
CN1523518A (zh) * 2003-02-17 2004-08-25 郭慧民 智能汉语文化辞典系统
JP4324058B2 (ja) * 2004-08-31 2009-09-02 キヤノン株式会社 画像処理装置及びその方法
GB2419764B (en) * 2004-11-01 2010-06-09 Sony Uk Ltd Encoding and detecting apparatus
JP2006135596A (ja) * 2004-11-05 2006-05-25 Fuji Xerox Co Ltd 符号化装置、復号化装置、データファイル、符号化方法、復号化方法及びこれらのプログラム
JP4510092B2 (ja) * 2005-10-25 2010-07-21 富士通株式会社 電子透かしの埋め込み及び検出
CN100552603C (zh) * 2006-03-05 2009-10-21 刘国桢 汉语字词全息编码计算机手机输入方法及键盘
CN101098455B (zh) * 2006-06-28 2011-06-29 北京爱国者妙笔数码科技有限责任公司 利用读取的影像编码实现控制的点读装置
CN101098454B (zh) * 2006-06-28 2010-08-18 北京爱国者妙笔数码科技有限责任公司 利用读取的影像编码实现控制的点播系统
CN101122916A (zh) * 2007-09-17 2008-02-13 张卓 利用汉字编码的信息查询方法
SG155791A1 (en) * 2008-03-18 2009-10-29 Radiantrust Pte Ltd Method for embedding covert data in a text document using character rotation
US8630444B2 (en) * 2009-12-30 2014-01-14 Mitsubishi Electric Research Laboratories, Inc. Method for embedding messages into structure shapes
CN105074731B (zh) * 2012-12-19 2019-06-28 电装波动株式会社 信息码、信息码生成方法、信息码读取装置以及信息码应用系统
JP6252150B2 (ja) * 2013-03-27 2017-12-27 株式会社デンソーウェーブ 情報コード生成方法、情報コード、情報コード読取装置、及び情報コード利用システム
JP6167956B2 (ja) * 2013-09-20 2017-07-26 株式会社デンソーウェーブ 情報コードの生成方法、情報コード、情報コード読取装置、及び情報コード利用システム
CN104867022A (zh) * 2015-06-02 2015-08-26 吴华明 一种基于可变防伪编码的防伪方法
CN104951984A (zh) * 2015-06-03 2015-09-30 吴华明 一种基于产品编码的交友方法
CN105550279A (zh) * 2015-12-10 2016-05-04 天津海量信息技术有限公司 基于视觉的列表页识别方法
CN107302645B (zh) * 2017-04-27 2019-08-16 珠海赛纳打印科技股份有限公司 一种图像处理装置及其图像处理方法
CN108255958B (zh) * 2017-12-21 2022-05-03 百度在线网络技术（北京）有限公司 数据查询方法、装置和存储介质
CN108765522B (zh) * 2018-05-15 2022-08-02 维沃移动通信有限公司 一种动态图像生成方法及移动终端
EP3763871A1 (en) 2019-07-08 2021-01-13 Comelz S.p.A. Accessory apparatus for feeding sheets of material to be cut in numerically controlled machines (ncms), and ncm comprising said accessory apparatus

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101763516A (zh) * 2010-01-15 2010-06-30 南京航空航天大学 一种基于拟合函数的文字识别方法
CN101908290A (zh) * 2010-07-27 2010-12-08 李水超 英语读物及其识读机
CN102902968A (zh) * 2011-07-25 2013-01-30 上海博路信息技术有限公司 一种手机扫描快速获取出版物内容的方法
US20170090693A1 (en) * 2015-09-25 2017-03-30 Lg Electronics Inc. Mobile terminal and method of controlling the same
CN108830126A (zh) * 2018-06-20 2018-11-16 上海凌脉网络科技股份有限公司 一种基于图像智能识别的产品营销互动方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3913536A4 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113343639A (zh) * 2021-05-19 2021-09-03 网易（杭州）网络有限公司 产品标识码图生成、基于产品标识码图的信息查询方法
CN113343639B (zh) * 2021-05-19 2023-10-03 网易（杭州）网络有限公司 产品标识码图生成、基于产品标识码图的信息查询方法

Also Published As

Publication number Publication date
EP3913536A4 (en) 2022-03-23
EP3913536A1 (en) 2021-11-24
CN109766978B (zh) 2020-06-16
CN109766978A (zh) 2019-05-17
US20220067469A1 (en) 2022-03-03
JP7130881B2 (ja) 2022-09-05
US11334780B2 (en) 2022-05-17
JP2022523651A (ja) 2022-04-26

Similar Documents

Publication Publication Date Title
US20170011732A1 (en) Low-vision reading vision assisting system based on ocr and tts
JP2002526862A (ja) ドキュメントを表わすデータの操作および表示のための他のフォーマットへの変換
WO2021121158A1 (zh) 公文文件处理方法、装置、计算机设备及存储介质
CN114547274B (zh) 多轮问答的方法、装置及设备
TW200416583A (en) Definition data generation method of account book voucher and processing device of account book voucher
TW201316187A (zh) 偵測及校正中文錯字的系統及方法
JP2011141749A (ja) 文書画像生成装置、文書画像生成方法及びコンピュータプログラム
JPH08147446A (ja) 電子ファイリング装置
CN103136453A (zh) 文档操作题的自动组卷方法和自动阅卷方法
WO2020147140A1 (zh) 一种词码的生成方法、识别方法、装置、存储介质
CN110991303A (zh) 一种图像中文本定位方法、装置及电子设备
CN113255331B (zh) 文本纠错方法、装置及存储介质
Thammarak et al. Automated data digitization system for vehicle registration certificates using google cloud vision API
CN110516125B (zh) 识别异常字符串的方法、装置、设备及可读存储介质
CN116225956A (zh) 自动化测试方法、装置、计算机设备和存储介质
US20220138416A1 (en) Dictionary editing apparatus, dictionary editing method, and recording medium recording thereon dictionary editing program
JP2019175037A (ja) 文字認識装置、方法およびプログラム
JPH0388062A (ja) 文書作成装置
CN111444716A (zh) 标题分词方法、终端及计算机可读存储介质
JP5604276B2 (ja) 文書画像生成装置および文書画像生成方法
JP2020166658A (ja) 情報処理装置、情報処理方法及びプログラム
US11170182B2 (en) Braille editing method using error output function, recording medium storing program for executing same, and computer program stored in recording medium for executing same
US8972239B2 (en) Syntax analysis information generation apparatus, translation apparatus, translation system, syntax analysis information generating method and computer program
WO2023200475A1 (en) Automatic form filling based on decoded information from machine-readable identifier
JP2001022773A (ja) イメージ文書のキーワード抽出方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19910826

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2021541706

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2019910826

Country of ref document: EP

Effective date: 20210817