CN108875727A - The detection method and device of graph-text identification, storage medium, processor - Google Patents
The detection method and device of graph-text identification, storage medium, processor Download PDFInfo
- Publication number
- CN108875727A CN108875727A CN201810712827.3A CN201810712827A CN108875727A CN 108875727 A CN108875727 A CN 108875727A CN 201810712827 A CN201810712827 A CN 201810712827A CN 108875727 A CN108875727 A CN 108875727A
- Authority
- CN
- China
- Prior art keywords
- graph
- text identification
- similarity
- character
- sample
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
This application discloses a kind of detection method and device of graph-text identification, storage medium, processors.Wherein, this method includes：Extract the first character and the first figure in graph-text identification to be detected；Determine the first similarity of the second character and the second similarity of the first figure and second graph in sample graph-text identification in the first character and at least one sample graph-text identification；First object sample graph-text identification and the second target sample graph-text identification are selected from least one sample graph-text identification respectively according to the first similarity and the second similarity, and using first object sample graph-text identification and the second target sample graph-text identification as testing result；Output test result.When present application addresses carrying out similar brand detection, due to handling using entire trade mark as piece image, similar brand caused by its separately processing inaccurate technical problem is not detected into according to feature.
Description
Technical field
This application involves graph text information searching field, in particular to a kind of graph-text identification detection method and device,
Storage medium, processor.
Background technique
Trade mark is can to distinguish the commodity or service that the commodity of oneself enterprise or service are provided with other enterprises
Mark.With China's economic development, the implementation of the expansion of reform and opening-up and Trademark Strategy, trade mark registration applications are year by year significantly
Increase.As trade mark examination & approval are getting faster, the quantity of application is more and more, and people acquaint oneself of the brand for establishing oneself, still
The application difficulty of trade mark is also increasing, and more and more trade marks are because have higher similarity with the trade mark applied for the registration of
It is rejected.
With online infringement, network sell-fake-products problem it is increasing, enterprise to trade mark carry out network monitoring also become more next
It is more necessary.On the one hand, enterprise can be carried out periodic retrieval to enterprise trademark, be obtained in time by State Administration for Industry & Commerce's trade mark board web
Know trade mark Anomalous dynamics, and checks whether that other people register similar mark or the behavior in other Category Accreditations identic trade marks, with
Just enterprise can quickly take counter-measure.On the other hand, enterprise needs in major electric business platform and sells the professional of product
Website carries out key monitoring, grasps network sell-fake-products information comprehensively, saves relevant evidence, is complained in time to each large platform, will
The effectively generation of containment online infringement problem.
In the prior art carry out similar brand detection when, be to be handled entire trade mark as a sub-picture, not according to
Feature separately handles it, and the accuracy for causing similar brand to detect is not high.
For above-mentioned problem, currently no effective solution has been proposed.
Summary of the invention
The embodiment of the present application provides a kind of detection method and device of graph-text identification, storage medium, processor, at least
When solving to carry out similar brand detection, due to being handled using entire trade mark as piece image, do not separated according to feature
Similar brand caused by processing detects inaccurate technical problem.
According to the one aspect of the embodiment of the present application, a kind of detection method of graph-text identification is provided, including：
Extract the first character and the first figure in graph-text identification to be detected；Determine the first character and at least one sample graph
The first similarity of the second character and the first figure are similar to second of second graph in sample graph-text identification in text mark
Degree；First object sample picture and text are selected from least one sample graph-text identification respectively according to the first similarity and the second similarity
Mark and the second target sample graph-text identification, and using first object sample graph-text identification and the second target sample graph-text identification as
Testing result；Output test result.
Optionally, first is selected from least one sample graph-text identification respectively according to the first similarity and the second similarity
Target sample graph-text identification and the second target sample graph-text identification, including：By at least one sample graph-text identification according to the first phase
It is ranked up like the sequence of degree from big to small, obtains first sample list；By M sample picture and text mark preceding in the first sample list
Know and is used as first object sample graph-text identification；By at least one sequence of the sample graph-text identification according to the second similarity from big to small
It is ranked up, obtains the second sample list；Using top n sample graph-text identification in second sample list as the second target sample
Graph-text identification.
Optionally, first is selected from least one sample graph-text identification respectively according to the first similarity and the second similarity
Target sample graph-text identification and the second target sample graph-text identification, including：Compare the first similarity and first threshold；Determination is greater than
Sample graph-text identification corresponding to first similarity of first threshold is as first object sample graph-text identification；It is similar to compare second
Degree and second threshold；Sample graph-text identification corresponding to the second similarity of second threshold be will be greater than as the second target sample figure
Text mark.
Optionally it is determined that the first similarity of the first character and the second character at least one sample graph-text identification, and
In first figure and sample graph-text identification before the second similarity of second graph, method further includes：Extract at least one sample
Character and figure in graph-text identification, obtain the second character and second graph.
Optionally it is determined that the first similarity of the first character and the second character at least one sample graph-text identification, including：
The minimum unit feature of the first character and the second character is determined respectively；Vectorization processing is carried out to minimum unit feature, is obtained most
The vector of junior unit feature；The similarity that the first character and the second character are determined using vector, by the first determining character and
The similarity of two characters is as the first similarity.
Optionally, the first character and the second character include：English, number and Chinese character；The first character and the are determined respectively
The minimum unit feature of two characters includes：It is minimum that English, number and Chinese character are expressed as English minimal characteristic unit, number
Feature unit and Chinese character minimal characteristic unit, wherein English minimal characteristic unit is independent letter in 26 English alphabets, number
Word minimal characteristic unit is independent number in 0 to 9, and Chinese character minimal characteristic unit is the list that can indicate graph-text identification feature
Word；Vectorization processing is carried out to minimum unit feature, obtains the vector of minimum unit feature, including：By 26 English alphabets and 0
It is 36 binary vectors to digital representation independent in 9, only one value is 1 in binary vector, remaining is all 0；By the Chinese
Word minimal characteristic cell attribute is real number value vector.
Optionally it is determined that the second similarity of the first figure and second graph at least one sample graph-text identification, including：
Targeted graphical is subjected to gray processing processing, wherein targeted graphical includes：First figure and second graph；Mesh after calculating gray processing
The average gray value for shape of marking on a map；The gray value and average gray value for comparing each pixel of figure after gray processing, if pixel
Gray value be greater than average gray value and be then denoted as 1, it is on the contrary then be denoted as 0, and be arranged in binary fingerprints by preset order and encode, obtain
To and corresponding first finger-print codes of the first figure and the second finger-print codes corresponding with second graph；Compare the first finger-print codes
With the second finger-print codes, identical data bits in the first finger-print codes and the second finger-print codes is determined, wherein the data bits
For reflecting the second similarity.
According to the another aspect of the embodiment of the present application, a kind of detection device of graph-text identification is additionally provided, including：Extract mould
Block, for extracting the first character and the first figure in graph-text identification to be detected；Determining module, for determine the first character with extremely
First similarity of the second character and the first figure and second graph in sample graph-text identification in a few sample graph-text identification
The second similarity；Selecting module, for the first similarity of foundation and the second similarity respectively from least one sample picture and text mark
Select first object sample graph-text identification and the second target sample graph-text identification in knowledge, and by first object sample graph-text identification and
Second target sample graph-text identification is as testing result；Output module is used for output test result.
According to the another aspect of the embodiment of the present application, a kind of storage medium is additionally provided, storage medium includes the journey of storage
Sequence, wherein the detection method of the graph-text identification in program operation where control storage medium more than equipment execution.
According to the another aspect of the embodiment of the present application, a kind of processor is additionally provided, processor is used to run program,
In, the detection method of graph-text identification when program is run more than execution.
In the embodiment of the present application, using the first character and the first figure extracted in graph-text identification to be detected；Determine
First similarity of the second character and the first figure and sample graph-text identification in one character and at least one sample graph-text identification
Second similarity of middle second graph；According to the first similarity and the second similarity respectively from least one sample graph-text identification
Select first object sample graph-text identification and the second target sample graph-text identification, and by first object sample graph-text identification and second
Target sample graph-text identification is as testing result；The mode of output test result, by the detection for carrying out similar graph-text identification
When, by graph-text identification figure and character separately handle, achieved the purpose that the similar picture and text label detection accuracy of raising, from
And realize the technical effect for more accurately detecting similar graph-text identification, and then solve carry out similar brand detection when, due to
It handles, does not detect similar brand caused by its separately processing according to feature inaccurate using entire trade mark as piece image
The technical issues of.
Detailed description of the invention
The drawings described herein are used to provide a further understanding of the present application, constitutes part of this application, this Shen
Illustrative embodiments and their description please are not constituted an undue limitation on the present application for explaining the application.In the accompanying drawings：
Fig. 1 is the flow chart according to a kind of detection method of graph-text identification of the embodiment of the present application；
Fig. 2 is the flow chart of the processing method of figure in a kind of graph-text identification according to the embodiment of the present application；
Fig. 3 is the flow chart of the similarity calculating method of figure in a kind of graph-text identification according to the embodiment of the present application；
Fig. 4 is the process of the average gray value calculating method of figure in a kind of graph-text identification according to the embodiment of the present application
Figure；
Fig. 5 is the structure chart according to a kind of detection device of graph-text identification of the embodiment of the present application.
Specific embodiment
In order to make those skilled in the art more fully understand application scheme, below in conjunction in the embodiment of the present application
Attached drawing, the technical scheme in the embodiment of the application is clearly and completely described, it is clear that described embodiment is only
The embodiment of the application a part, instead of all the embodiments.Based on the embodiment in the application, ordinary skill people
Member's every other embodiment obtained without making creative work, all should belong to the model of the application protection
It encloses.
It should be noted that the description and claims of this application and term " first " in above-mentioned attached drawing, "
Two " etc. be to be used to distinguish similar objects, without being used to describe a particular order or precedence order.It should be understood that using in this way
Data be interchangeable under appropriate circumstances, so as to embodiments herein described herein can in addition to illustrating herein or
Sequence other than those of description is implemented.In addition, term " includes " and " having " and their any deformation, it is intended that cover
Cover it is non-exclusive include, for example, the process, method, system, product or equipment for containing a series of steps or units are not necessarily limited to
Step or unit those of is clearly listed, but may include be not clearly listed or for these process, methods, product
Or other step or units that equipment is intrinsic.
According to the embodiment of the present application, a kind of detection method embodiment of graph-text identification is provided, it should be noted that attached
The step of process of figure illustrates can execute in a computer system such as a set of computer executable instructions, though also,
So logical order is shown in flow charts, but in some cases, it can be to be different from shown by sequence execution herein
Or the step of description.
Fig. 1 is according to a kind of flow chart of the detection method of graph-text identification of the embodiment of the present application, as shown in Figure 1, the party
Method includes the following steps：
Step S102 extracts the first character and the first figure in graph-text identification to be detected.
In some embodiments of the present application, by taking trade mark as an example：The detection of trade mark similarity mainly to two trade marks it
Between the similarity degree of content give a mark, the close degree of branding content is judged according to the height of score.Due to trade mark be by
The connected regions such as text, figure, number, letter (including capital and small letter) composition, and the color and intensity of every piece of connected region
It is almost the same, that is to say, that these regions of trade mark may be considered relatively significantly, therefore first to quotient in the embodiment of the present application
Figure, letter, number, Chinese character in mark extract, and then analyze each region.
Above-mentioned first character includes at least one of letter, number, Chinese character.Pass through the key character to trade mark to be detected
Extracted respectively with figure, thus by trade mark figure and character separately handle, can more accurately detect similar
Trade mark.
Step S104 determines the first similarity of the second character in the first character and at least one sample graph-text identification, with
And first second graph in figure and sample graph-text identification the second similarity.
There are many methods of determination of first similarity, such as：Phase is calculated using Euclidean distance calculating formula of similarity
Like degree, manhatton distance calculation formula calculates similarity, cosine similarity formula etc..An optional implementation as the application
Example calculates the first similarity by cosine similarity formula, determines especially by following manner, but not limited to this：It determines respectively
The minimum unit feature of first character and the second character；Vectorization processing is carried out to minimum unit feature, obtains minimum unit spy
The vector of sign；The similarity that the first character and the second character are determined using vector, by determining the first character and the second character
Similarity is as the first similarity.
In some alternative embodiments of the application, above-mentioned minimal characteristic unit be may include one or more character elements
, for example, can also be incited somebody to action using a letter in 26 English alphabets as minimal characteristic unit for English character
In 26 English alphabets two (or more) letter be used as minimal characteristic unit, wherein for the latter, although accuracy is opposite
It is relatively low in the former, however, it is possible to improve computational efficiency.
Specifically, the minimum unit feature of the first character and the second character can be determined in the following manner：By English, number
Word and Chinese character are expressed as English minimal characteristic unit, digital minimal characteristic unit and Chinese character minimal characteristic unit, wherein English
Literary minimal characteristic unit is independent letter in 26 English alphabets, and digital minimal characteristic unit is independent number in 0 to 9,
Chinese character minimal characteristic unit is the word that can indicate graph-text identification feature；
As the alternative embodiment of the application, the vector of minimum unit feature can determine in the following manner, but
It is without being limited thereto：It is 36 binary vectors by independent digital representation in 26 English alphabets and 0 to 9, in binary vector only
Having a value is 1, remaining is all 0, such as letter a (or A) can be expressed as [1,0,0 ..., 0] for Chinese character minimal characteristic list
Member is characterized as real number value vector, and as the alternative embodiment of the application, Chinese character word is characterized as real number value using Word2vec
Vector, Word2vec are the correlation models for generating term vector.
In some alternative embodiments of the application, the sample from the vector sum database for the character that trade mark to be detected extracts is utilized
The vector of the character of this trade mark carries out similarity calculation using following formula：
Wherein, vector A=(A1, A2 ..., An), B=(B1, B2 ..., Bn), vector A are character in trade mark to be detected
Vector, B be sample trade mark in character vector.
As the alternative embodiment of the application, the second figure in the first figure and at least one sample graph-text identification is determined
Second similarity of shape, including：Targeted graphical is subjected to gray processing processing, wherein targeted graphical includes：First figure and second
Figure；The average gray value of targeted graphical after calculating gray processing；Compare after gray processing the gray value of each pixel of figure and flat
On the contrary equal gray value is denoted as 1 if the gray value of pixel is greater than average gray value, then be denoted as 0, and by preset order arrangement
It is encoded at binary fingerprints, obtains and corresponding first finger-print codes of the first figure and the second fingerprint corresponding with second graph and compile
Code；Compare the first finger-print codes and the second finger-print codes, determines identical data in the first finger-print codes and the second finger-print codes
Digit, wherein the data bits is for reflecting the second similarity.
Fig. 2 is the flow chart of the processing method of figure in a kind of graph-text identification according to the embodiment of the present application, such as Fig. 2 institute
Show, includes the following steps：
Step S202 extracts the targeted graphical in trade mark.To the sample trade mark in trade mark to be detected and database
In targeted graphical extract.
Feature size is normalized to the size of 8*8 by step S204.Figure normalization, which refers to, carries out a series of marks to figure
Quasi- processing transformation is allowed to be transformed to the process of a fixed standard form, and the test pattern is at referred to as normalization figure.Figure warp
Cross characteristic block cutting after, the characteristic block for cutting generation is not of uniform size, at this moment it is necessary to take it is normalized operation come uniform characteristics block
The size of figure, the size of above-mentioned 8*8 refer to the figure that figure is normalized to 8*8 pixel.
Step S206 carries out image enhancement to the figure after normalization.Figure enhancing refers to the useful letter in enhancing figure
Breath, it can be the process of a distortion, it is therefore an objective to improve the visual effect of figure.
Step S208 carries out gray processing to enhanced figure.Figure gray processing, which refers to, allows each of figure pixel
Point all meets lower relation of plane：R=G=B (R is the value of red variable, and G is the value of green variable, and B is the value of blue variable, this
Three values are equal and are an integer less than 255), this value at this time is called gray value.
Step S210 carries out shape similarity calculating.
Step S212 exports highest 3 similar marks of similarity.
Fig. 3 is the flow chart of the similarity calculating method of figure in a kind of graph-text identification according to the embodiment of the present application.Step
Shape similarity calculating specifically includes following steps in rapid S210, as shown in Figure 3：
Step S302 calculates the average gray value of figure.
Step S304 compares the gray value and average gray value of each pixel in figure.If the gray value of pixel is greater than figure
The average gray value of shape is then denoted as 1, small, is denoted as 0, is arranged in the finger-print codes of 64 2 systems in a certain order.
Step S306, the finger-print codes of trade mark and sample brand logo more to be detected.How many position in 64 calculated
It is different, if different data bits is no more than 5, just illustrates that two pictures are much like, if it is greater than 10, illustrate it
Be two different pictures.
Fig. 4 is the process of the average gray value calculating method of figure in a kind of graph-text identification according to the embodiment of the present application
Figure, the fall into a trap average gray value of nomogram shape of step S302 specifically include following steps, as shown in Figure 4：
Step S402 traverses the pixel of figure to be processed.
Step S404 carries out cumulative summation to the gray value of each pixel.Wherein, the gray value of pixel mainly calculates
Method is as follows：Any color is all made of Red Green Blue, if the color of certain original point is that (R represents red to RGB, G
Green is represented, B represents blue), the gray value of some pixel in figure can be calculated by following formula：
Gray=0.3*R+0.59*G+0.11*B
Wherein, Gray is the gray value of pixel, R=G=B and be integer less than 255.
Step S406 calculates the pixel total number of figure.
Step S408 calculates the average gray value of figure.
The average gray value of figure is obtained with being divided by with pixel total number for the pixel gray value being calculated.
Based on step S104, can be utilized respectively the sample trade mark in the trade mark and database to be detected of extraction character and
Figure carries out similarity calculation, determines the first similarity of the character in trade mark and database to be detected in sample trade mark, according to
The similarity finds out sample trade mark similar with trade mark to be detected in database；Determine sample quotient in trade mark and database to be detected
Second similarity of the figure in mark finds out sample trade mark similar with trade mark to be detected in database according to the similarity.
Step S106 selects the according to the first similarity and the second similarity from least one sample graph-text identification respectively
One target sample graph-text identification and the second target sample graph-text identification, and by first object sample graph-text identification and the second target sample
This graph-text identification is used as testing result.
As the alternative embodiment of the application, according to the first similarity and the second similarity respectively from least one sample
First object sample graph-text identification and the second target sample graph-text identification are selected in this graph-text identification, including：By at least one sample
This graph-text identification is ranked up according to the sequence of the first similarity from big to small, obtains first sample list；By the first sample
Preceding M sample graph-text identification is as first object sample graph-text identification in list.Using extraction trade mark to be detected character and
The character of at least one sample trade mark in database carries out similarity calculation, and character similarity can be characterized by obtaining at least one
Size as a result, by the result according to from big to small sequence sort, form a similar sample trade mark list, the application's
In one alternative embodiment, similar brand of the preceding 3 sample trade marks as trade mark to be detected is taken from above-mentioned sample trade mark list,
As testing result.
As the alternative embodiment of the application, according to the first similarity and the second similarity respectively from least one sample graph
First object sample graph-text identification and the second target sample graph-text identification are selected in text mark, including：By at least one sample graph
Text mark is ranked up according to the sequence of the second similarity from big to small, obtains the second sample list；By second sample list
Middle top n sample graph-text identification is as the second target sample graph-text identification.Utilize the figure and data of the trade mark to be detected of extraction
The figure of at least one sample trade mark in library carries out similarity calculation, and shape similarity size can be characterized by obtaining at least one
As a result, by the result according to from big to small sequence sort, form a similar sample trade mark list, at one of the application
In alternative embodiment, similar brand of the preceding 3 sample trade marks as trade mark to be detected is taken from above-mentioned sample trade mark list, as
Testing result.
As another alternative embodiment of the application, according to the first similarity and the second similarity respectively from least one
First object sample graph-text identification and the second target sample graph-text identification are selected in sample graph-text identification, including：Compare the first phase
Like degree and first threshold；It determines and is greater than sample graph-text identification corresponding to the first similarity of first threshold as first object sample
This graph-text identification；Compare the second similarity and second threshold；It will be greater than sample graph corresponding to the second similarity of second threshold
Text mark is used as the second target sample graph-text identification.In some alternative embodiments of the application, a word is preset respectively
The threshold value of similarity and shape similarity is accorded with as benchmark, utilizes the value and image similarity of practical calculated character similarity
Value and preset threshold value be compared, if the value of practical calculated similarity is greater than preset threshold value,
Determine that the corresponding sample trade mark of the similarity calculation result is similar to trade mark to be detected.
As the alternative embodiment of the application, the second word in the first character and at least one sample graph-text identification is determined
In first similarity of symbol and the first figure and sample graph-text identification before the second similarity of second graph, method is also wrapped
It includes：The character and figure at least one sample graph-text identification are extracted, the second character and second graph are obtained.To quotient to be detected
When the extraction of mark progress character and figure, while the character of sample trade mark and figure in database are extracted, it can also be pre-
The character and figure of sample trade mark in database are first extracted, when there is new sample trade mark that database is added, timely update data
Library, to guarantee the accuracy of similar brand detection.
Step S108, output test result.
3 sample trade mark similar with trade mark to be detected will be obtained, and will be similar by figure by character similarity calculation
Degree calculates, and 3 sample trade mark similar with trade mark to be detected is obtained, totally 6 similar sample trade marks, as testing result.
Through the above steps, may be implemented by trade mark figure and character separately handle, can be improved so similar
The accuracy of trade mark detection.
Fig. 5 is according to a kind of structure chart of the detection device of graph-text identification of the embodiment of the present application, as shown in figure 5, including：
Extraction module 50, for extracting the first character and the first figure in graph-text identification to be detected；Determining module 52, for determining
First similarity of the second character and the first figure and sample graph-text identification in one character and at least one sample graph-text identification
Second similarity of middle second graph；Selecting module 54, for the first similarity of foundation and the second similarity respectively from least one
Select first object sample graph-text identification and the second target sample graph-text identification in a sample graph-text identification, and by first object sample
This graph-text identification and the second target sample graph-text identification are as testing result；Output module 56 is used for output test result.
It should be noted that the correlation that the preferred embodiment of embodiment illustrated in fig. 5 may refer to embodiment illustrated in fig. 1 is retouched
It states, details are not described herein again.
The embodiment of the present application also provides a kind of storage medium, storage medium includes the program of storage, wherein is transported in program
Equipment where controlling storage medium when row executes the detection method of the above graph-text identification.
Above-mentioned storage medium is used to store the program for executing following functions：Extract the first character in graph-text identification to be detected
With the first figure；Determine the first similarity and first of the second character in the first character and at least one sample graph-text identification
Second similarity of second graph in figure and sample graph-text identification；According to the first similarity and the second similarity respectively from least
Select first object sample graph-text identification and the second target sample graph-text identification in one sample graph-text identification, and by first object
Sample graph-text identification and the second target sample graph-text identification are as testing result；Output test result.
The embodiment of the present application also provides a kind of processor, processor is for running program, wherein holds in program operation
The detection method of the above graph-text identification of row.
Above-mentioned processor is used to execute the program for realizing following functions：Extract the first character in graph-text identification to be detected and
First figure；Determine the first similarity and the first figure of the second character in the first character and at least one sample graph-text identification
Second similarity of second graph in shape and sample graph-text identification；According to the first similarity and the second similarity respectively from least one
Select first object sample graph-text identification and the second target sample graph-text identification in a sample graph-text identification, and by first object sample
This graph-text identification and the second target sample graph-text identification are as testing result；Output test result.
Above-mentioned the embodiment of the present application serial number is for illustration only, does not represent the advantages or disadvantages of the embodiments.
In above-described embodiment of the application, all emphasizes particularly on different fields to the description of each embodiment, do not have in some embodiment
The part of detailed description, reference can be made to the related descriptions of other embodiments.
In several embodiments provided herein, it should be understood that disclosed technology contents can pass through others
Mode is realized.Wherein, the apparatus embodiments described above are merely exemplary, such as the division of the unit, Ke Yiwei
A kind of logical function partition, there may be another division manner in actual implementation, for example, multiple units or components can combine or
Person is desirably integrated into another system, or some features can be ignored or not executed.Another point, shown or discussed is mutual
Between coupling, direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING or communication link of unit or module
It connects, can be electrical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
On unit.It can some or all of the units may be selected to achieve the purpose of the solution of this embodiment according to the actual needs.
It, can also be in addition, each functional unit in each embodiment of the application can integrate in one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, the technical solution of the application is substantially
The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words
It embodies, which is stored in a storage medium, including some instructions are used so that a computer
Equipment (can for personal computer, server or network equipment etc.) execute each embodiment the method for the application whole or
Part steps.And storage medium above-mentioned includes：USB flash disk, read-only memory (ROM, Read-Only Memory), arbitrary access are deposited
Reservoir (RAM, Random Access Memory), mobile hard disk, magnetic or disk etc. be various to can store program code
Medium.
The above is only the preferred embodiment of the application, it is noted that for the ordinary skill people of the art
For member, under the premise of not departing from the application principle, several improvements and modifications can also be made, these improvements and modifications are also answered
It is considered as the protection scope of the application.
Claims (10)
1. a kind of detection method of graph-text identification, which is characterized in that including：
Extract the first character and the first figure in graph-text identification to be detected；
Determine the first similarity and described first of the second character in first character and at least one sample graph-text identification
Second similarity of second graph in figure and the sample graph-text identification；
Select from least one described sample graph-text identification respectively according to first similarity and second similarity
One target sample graph-text identification and the second target sample graph-text identification, and by the first object sample graph-text identification and described
Two target sample graph-text identifications are as testing result；
Export the testing result.
2. the method according to claim 1, wherein according to first similarity and second similarity point
First object sample graph-text identification and the second target sample graph-text identification are selected not from least one described sample graph-text identification,
Including：
At least one described sample graph-text identification is ranked up according to the sequence of first similarity from big to small, obtains
One sample list；Using M sample graph-text identification preceding in the first sample list as the first object sample graph-text identification；
At least one described sample graph-text identification is ranked up according to the sequence of second similarity from big to small, obtains
Two sample lists；Using top n sample graph-text identification in second sample list as the second target sample graph-text identification.
3. the method according to claim 1, wherein according to first similarity and second similarity point
First object sample graph-text identification and the second target sample graph-text identification are selected not from least one described sample graph-text identification,
Including：
Compare first similarity and first threshold；It determines and is greater than corresponding to first similarity of the first threshold
Sample graph-text identification is as the first object sample graph-text identification；
Compare second similarity and second threshold；It will be greater than sample corresponding to second similarity of the second threshold
This graph-text identification is as the second target sample graph-text identification.
4. the method according to claim 1, wherein determining first character and at least one sample picture and text mark
First similarity of the second character and the second phase of first figure and second graph in the sample graph-text identification in knowledge
Before degree, the method also includes：
The character and figure at least one described sample graph-text identification are extracted, second character and second graph are obtained.
5. the method according to claim 1, wherein determining first character and at least one sample picture and text mark
First similarity of the second character in knowledge, including：
The minimum unit feature of first character and second character is determined respectively；To the minimum unit feature carry out to
Quantification treatment obtains the vector of minimum unit feature；First character and second character are determined using the vector
Similarity, using the similarity of determining first character and second character as first similarity.
6. according to the method described in claim 5, it is characterized in that, first character and second character include：English
Text, number and Chinese character；
The minimum unit feature for determining first character and the second character respectively includes：By the English, the number and institute
It states Chinese character and is expressed as English minimal characteristic unit, digital minimal characteristic unit and Chinese character minimal characteristic unit, wherein is described
English minimal characteristic unit is independent letter in 26 English alphabets, and the number minimal characteristic unit is independent in 0 to 9
Number, the Chinese character minimal characteristic unit are the word that can indicate the graph-text identification feature；
Vectorization processing is carried out to the minimum unit feature, obtains the vector of minimum unit feature, including：By 26 English
Independent digital representation is 36 binary vectors in text mother and described 0 to 9, only one value is in the binary vector
1, remaining is all 0；It is real number value vector by the Chinese character minimal characteristic cell attribute.
7. the method according to claim 1, wherein determining first figure and at least one sample picture and text mark
Second similarity of second graph in knowledge, including：
Targeted graphical is subjected to gray processing processing, wherein the targeted graphical includes：First figure and second graph；Meter
The average gray value of the targeted graphical after calculation gray processing；Compare after gray processing the gray value of each pixel of figure and flat
On the contrary equal gray value is denoted as 1 if the gray value of the pixel is greater than the average gray value, then be denoted as 0, and by presetting
Sequence be arranged in binary fingerprints coding, obtain the first finger-print codes corresponding with first figure and with the second graph
Corresponding second finger-print codes；Compare first finger-print codes and the second finger-print codes, determine first finger-print codes and
Identical data bits in second finger-print codes, wherein the data bits is for reflecting second similarity.
8. a kind of detection device of graph-text identification, which is characterized in that including：
Extraction module, for extracting the first character and the first figure in graph-text identification to be detected；
Determining module, for determining that first character is similar to first of the second character at least one sample graph-text identification
Second similarity of second graph in degree and first figure and the sample graph-text identification；
Selecting module is used for according to first similarity and second similarity respectively from least one described sample picture and text
Select first object sample graph-text identification and the second target sample graph-text identification in mark, and by the first object sample picture and text
Mark and the second target sample graph-text identification are as testing result；
Output module, for exporting the testing result.
9. a kind of storage medium, which is characterized in that the storage medium includes the program of storage, wherein described program right of execution
Benefit require any one of 1 to 7 described in graph-text identification detection method.
10. a kind of processor, which is characterized in that the processor is for running program, wherein right of execution when described program is run
Benefit require any one of 1 to 7 described in graph-text identification detection method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810712827.3A CN108875727B (en) | 2018-06-29 | 2018-06-29 | The detection method and device of graph-text identification, storage medium, processor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810712827.3A CN108875727B (en) | 2018-06-29 | 2018-06-29 | The detection method and device of graph-text identification, storage medium, processor |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108875727A true CN108875727A (en) | 2018-11-23 |
| CN108875727B CN108875727B (en) | 2019-08-30 |
Family
ID=64296610
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810712827.3A Active CN108875727B (en) | 2018-06-29 | 2018-06-29 | The detection method and device of graph-text identification, storage medium, processor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108875727B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110929720A (en) * | 2019-10-28 | 2020-03-27 | 杭州电子科技大学 | Component detection method based on LOGO matching and OCR |
| CN111259207A (en) * | 2018-11-30 | 2020-06-09 | 阿里巴巴集团控股有限公司 | Short message identification method, device and equipment |
| CN112037088A (en) * | 2020-11-02 | 2020-12-04 | 北京梦知网科技有限公司 | Trademark infringement determining method and device |
| CN112561389A (en) * | 2020-12-23 | 2021-03-26 | 北京元心科技有限公司 | Method and device for determining equipment detection result and electronic equipment |
| CN112861656A (en) * | 2021-01-21 | 2021-05-28 | 平安科技（深圳）有限公司 | Trademark similarity detection method and device, electronic equipment and storage medium |
| CN113780268A (en) * | 2020-12-28 | 2021-12-10 | 北京沃东天骏信息技术有限公司 | Trademark identification method and device and electronic equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103258037A (en) * | 2013-05-16 | 2013-08-21 | 西安工业大学 | Trademark identification searching method for multiple combined contents |
| CN104462380A (en) * | 2014-12-11 | 2015-03-25 | 北京中细软移动互联科技有限公司 | Trademark retrieval method |
| CN107239784A (en) * | 2017-07-03 | 2017-10-10 | 福建中金在线信息科技有限公司 | A kind of image identification method, device, electronic equipment and readable storage medium storing program for executing |
-
2018
- 2018-06-29 CN CN201810712827.3A patent/CN108875727B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103258037A (en) * | 2013-05-16 | 2013-08-21 | 西安工业大学 | Trademark identification searching method for multiple combined contents |
| CN104462380A (en) * | 2014-12-11 | 2015-03-25 | 北京中细软移动互联科技有限公司 | Trademark retrieval method |
| CN107239784A (en) * | 2017-07-03 | 2017-10-10 | 福建中金在线信息科技有限公司 | A kind of image identification method, device, electronic equipment and readable storage medium storing program for executing |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111259207A (en) * | 2018-11-30 | 2020-06-09 | 阿里巴巴集团控股有限公司 | Short message identification method, device and equipment |
| CN110929720A (en) * | 2019-10-28 | 2020-03-27 | 杭州电子科技大学 | Component detection method based on LOGO matching and OCR |
| CN110929720B (en) * | 2019-10-28 | 2022-11-04 | 杭州电子科技大学 | Component detection method based on LOGO matching and OCR |
| CN112037088A (en) * | 2020-11-02 | 2020-12-04 | 北京梦知网科技有限公司 | Trademark infringement determining method and device |
| CN112561389A (en) * | 2020-12-23 | 2021-03-26 | 北京元心科技有限公司 | Method and device for determining equipment detection result and electronic equipment |
| CN112561389B (en) * | 2020-12-23 | 2023-11-10 | 北京元心科技有限公司 | Method and device for determining detection result of equipment and electronic equipment |
| CN113780268A (en) * | 2020-12-28 | 2021-12-10 | 北京沃东天骏信息技术有限公司 | Trademark identification method and device and electronic equipment |
| CN112861656A (en) * | 2021-01-21 | 2021-05-28 | 平安科技（深圳）有限公司 | Trademark similarity detection method and device, electronic equipment and storage medium |
| CN112861656B (en) * | 2021-01-21 | 2024-05-14 | 平安科技（深圳）有限公司 | Trademark similarity detection method and device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108875727B (en) | 2019-08-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108875727B (en) | The detection method and device of graph-text identification, storage medium, processor | |
| Bi et al. | Multi-level dense descriptor and hierarchical feature matching for copy–move forgery detection | |
| Goel et al. | Dual branch convolutional neural network for copy move forgery detection | |
| CN112381775B (en) | Image tampering detection method, terminal device and storage medium | |
| CN113283446B (en) | Method and device for identifying object in image, electronic equipment and storage medium | |
| Zamri et al. | Tree species classification based on image analysis using Improved-Basic Gray Level Aura Matrix | |
| CN109740606B (en) | Image identification method and device | |
| Jaiprakash et al. | Low dimensional DCT and DWT feature based model for detection of image splicing and copy-move forgery | |
| CN111461164B (en) | Sample data set capacity expansion method and model training method | |
| Shabat et al. | Angled local directional pattern for texture analysis with an application to facial expression recognition | |
| CN113449725B (en) | Object classification method, device, equipment and storage medium | |
| Neelima et al. | Perceptual hash function based on scale-invariant feature transform and singular value decomposition | |
| CN112001362A (en) | Image analysis method, image analysis device and image analysis system | |
| CN111553241A (en) | Method, device and equipment for rejecting mismatching points of palm print and storage medium | |
| Nandanwar et al. | Forged text detection in video, scene, and document images | |
| JP2018013887A (en) | Feature selection device, tag relevant area extraction device, method, and program | |
| Bakshi et al. | ALPR-An Intelligent Approach Towards Detection and Recognition of License Plates in Uncontrolled Environments | |
| CN113609290A (en) | Address recognition method and device and storage medium | |
| Xiao et al. | Road marking detection based on structured learning | |
| Huang et al. | Chinese historic image threshold using adaptive K-means cluster and Bradley’s | |
| Singh et al. | Text Region Extraction: A Morphological Based Image Analysis Using Genetic Algorithm | |
| Oyiza et al. | An improved discrete cosine transformation block based scheme for copy-move image forgery detection | |
| CN117558011B (en) | Image text tampering detection method based on self-consistency matrix and multi-scale loss | |
| CN113111833B (en) | Safety detection method and device of artificial intelligence system and terminal equipment | |
| Xiao et al. | Combining HWEBING and HOG‐MLBP features for pedestrian detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP02 | Change in the address of a patent holder |
Address after: 519031 office 1316, No. 1, lianao Road, Hengqin new area, Zhuhai, Guangdong Patentee after: LONGMA ZHIXIN (ZHUHAI HENGQIN) TECHNOLOGY Co.,Ltd. Address before: 519000 room 417, building 20, creative Valley, Hengqin new area, Xiangzhou, Zhuhai, Guangdong Patentee before: LONGMA ZHIXIN (ZHUHAI HENGQIN) TECHNOLOGY Co.,Ltd. |
|
| CP02 | Change in the address of a patent holder |