CN108427959A - Board state collection method based on image recognition and system - Google Patents
Board state collection method based on image recognition and system Download PDFInfo
- Publication number
- CN108427959A CN108427959A CN201810124997.XA CN201810124997A CN108427959A CN 108427959 A CN108427959 A CN 108427959A CN 201810124997 A CN201810124997 A CN 201810124997A CN 108427959 A CN108427959 A CN 108427959A
- Authority
- CN
- China
- Prior art keywords
- image
- template
- recognition
- region
- identification
- 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.)
- Pending
Links
Classifications
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Library & Information Science (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Databases & Information Systems (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of board state collection method and system based on image recognition, and this approach includes the following steps：The image of board display is captured in real time；Image and the image template in template database are compared, and call the image template to match with image, the image received is divided into multiple identification regions according to the matched image template；Content recognition is carried out respectively for each identification region；The configuration information of identification region is converted into data and inserts monitoring data table, and the monitoring data table is sent to storage database.The present invention method and system can by board status information Formatting Output with facilitate monitoring and it is at low cost.
Description
Technical field
The present invention relates to data acquisition technology field more particularly to a kind of board state collection methods based on image recognition
And system.
Background technology
During printing or manufacturing shop include the intellectualized reconstruction of MES implementations, big data analysis or AI transformations, have not
The important state data of few board can not be acquired directly since genuine quotient does not provide data-interface etc..However, board state is believed
Breath is the basis of process parameter optimizing, product quality optimizing stability, and the acquisition of board status information is required.
Currently, the acquisition of board status information includes two ways：First, the installation data harvester on board leads to
It crosses and cracks agreement and come out board status information capture；Second, board status information capture is come out by board screen capture.
However, both the above mode has following defect：First way needs additional attachment means, it is also necessary to crack association
View, cost are too high；The second way is only capable of obtaining a large amount of picture, and required memory space is big while being inconvenient to monitor.
Therefore, it is necessary to it is a kind of can by board status information Formatting Output with facilitate monitoring and it is at low cost based on figure
As the board state collection method and system of identification.
Invention content
In view of the above problems, it is proposed that the present invention overcoming the above problem in order to provide one kind or solves at least partly
State the board state collection method and system based on image recognition of problem.
One aspect of the present invention provides a kind of board state collection method based on image recognition, including following step
Suddenly：
The image of board display is captured in real time；
Image and the image template in template database are compared, and call the image template to match with image,
The image received is divided into multiple identification regions according to the matched image template；
Content recognition is carried out respectively for each identification region；
The configuration information of identification region is converted into data and inserts monitoring data table, and the monitoring data table is sent to and is deposited
Store up database.
The board state collection method based on image recognition further includes：
The image template of board display is captured, and frame choosing is carried out to typical image region in the image template, and is defined
Data format and the storage side of mapping relations and the typical image region between the typical image region and monitoring data table
The image template defined is sent to template database and stored by formula.
The board state collection method based on image recognition further includes：The identified areas of specified image template.
It is described to compare image and the image template in template database, specifically include following steps：
With the presence or absence of the image template with images match in judge templet database；When in template database be not present and figure
When as matched image template, the process flow of None- identified is carried out；When there is image with images match in template database
When template, the content part of the identification region of image is deducted, to obtain the frame of image；By the frame of image and image template into
Row similarity evaluation；When, there are when identified areas, area being first identified between the frame and image template of image in image template
Domain similarity evaluation, then carry out overall similarity evaluation；When identified areas is not present in image template, the frame and figure of image
As directly carrying out overall similarity evaluation between template；The highest image template of frame entirety similarity of selection and image；When
When the overall similarity of the image template is higher than threshold value, the content recognition flow of image is carried out.
Similarity uses any in Euclidean distance algorithm, perceptual hash algorithm or feature point extraction algorithm.
It is described that the image received is divided into multiple identification regions according to matched image template, specifically include following step
Suddenly：
The definition information of parsing and the image template of images match；According to the definition information of image template, image is positioned
Identification region.
Another aspect of the present invention provides a kind of board state acquisition system based on image recognition, including：
Board display image harvester, the image for capturing board display in real time, and the image is sent
To template matches module；Template matches module, for receiving image in real time, by the image template in image and template database
It is compared, and calls the image template to match with image, divided the image received according to the matched image template
For multiple identification regions；Content recognition scheduler module is called for being directed to each identification region in the progress of content recognition engine respectively
Hold identification, is additionally operable to the monitoring data table being sent to storage database；Content recognition engine is used for the configuration of identification region
Information is converted to data filling monitoring data table, and the monitoring data table is sent to content recognition scheduler module.
The board display image harvester based on image recognition further includes：
Template definition is to guide module, and for receiving image template, frame choosing is carried out to typical image region in the image template,
And define mapping relations and the typical image region between the typical image region and monitoring data table data format and
The image template defined is sent to template database and stored by storage mode；Wherein, board display image acquisition dress
It sets, is additionally operable to the image template of capture board display, the image template is sent to template definition guide when board initializes
Module.
Board display image harvester is camera or video frequency collection card.
Identification region includes numerical identification region, text identification region, icon-based programming region, Table recognition region and curve
Identification region, content recognition engine include numerical value content identification engine, content of text identification engine, iconic content identification engine,
Table content identifies engine and curve content recognition engine.
The board state collection method based on image recognition and system of the present invention is by establishing template database come online
Matching will be divided into each identification region per frame board state picture, according to the matched image template of institute per frame board state picture,
And different types of content recognition engine is called to carry out content recognition to be directed to each identification region, to obtain number, word and song
The recognizable contents such as line point, not only shared memory space is small, but also convenient for monitoring, at low cost.
Above description is only the general introduction of technical solution of the present invention, in order to better understand the technical means of the present invention,
And can be implemented in accordance with the contents of the specification, and in order to allow above and other objects of the present invention, feature and advantage can
It is clearer and more comprehensible, below the special specific implementation mode for lifting the present invention.
Description of the drawings
By reading the detailed description of hereafter preferred embodiment, various other advantages and benefit are common for this field
Technical staff will become clear.Attached drawing only for the purpose of illustrating preferred embodiments, and is not considered as to the present invention
Limitation.And throughout the drawings, the same reference numbers will be used to refer to the same parts.In the accompanying drawings：
The step of Fig. 1 is a kind of board state collection method based on image recognition of the embodiment of the present invention is schemed；
Fig. 2 is the flow chart that the image template in the image and template database of the embodiment of the present invention is compared；
Fig. 3 is that a kind of board state acquisition system based on image recognition of the embodiment of the present invention connects block diagram.
Specific implementation mode
The exemplary embodiment of the disclosure is more fully described below with reference to accompanying drawings.Although showing the disclosure in attached drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here
It is limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure
Completely it is communicated to those skilled in the art.
Those skilled in the art of the present technique are appreciated that unless otherwise defined, all terms used herein (including technology art
Language and scientific terminology), there is meaning identical with the general understanding of the those of ordinary skill in fields of the present invention.Should also
Understand, those terms such as defined in the general dictionary, it should be understood that have in the context of the prior art
The consistent meaning of meaning, and unless by specific definitions, otherwise will not be explained with the meaning of idealization or too formal.
The step of Fig. 1 is a kind of board state collection method based on image recognition of the embodiment of the present invention is schemed, such as Fig. 1 institutes
Show, the board state collection method provided by the invention based on image recognition includes the following steps：
S1 captures the image of board display in real time.
S2 compares image and the image template in template database, and calls the image mould to match with image
The image received is divided into multiple identification regions by plate according to the matched image template.
Fig. 2 is the flow chart that the image template in the image and template database of the embodiment of the present invention is compared, such as Fig. 2
It is shown, it is described to compare the image and the image template in template database, specifically include following steps：
With the presence or absence of the image template with images match in judge templet database；When in template database be not present and figure
When as matched image template, the process flow of None- identified is carried out；When there is image with images match in template database
When template, the content part of the identification region of image is deducted, to obtain the frame of image；By the frame of image and image template into
Row similarity evaluation；When, there are when identified areas, area being first identified between the frame and image template of image in image template
Domain similarity evaluation, then carry out overall similarity evaluation；When identified areas is not present in image template, the frame and figure of image
As directly carrying out overall similarity evaluation between template；The highest image template of frame entirety similarity of selection and image；When
When the overall similarity of the image template is higher than threshold value, the content recognition flow of image is carried out.Threshold value can be according to image and figure
Set as the similarity history value of template, for example, similarity history value codomain ranging from 90%~100%, similarity history
The bigger value the more similar, then it is 90% that threshold value, which can be arranged,.
Wherein, it is raising computational efficiency, can preferentially judges whether preamble template similarity can receive, rather than every time
All study and judge all image templates.
Similarity uses any in Euclidean distance algorithm, perceptual hash algorithm or feature point extraction algorithm.
The core of Euclidean distance algorithm is：If image array has n element (n pixel), with n element value (x1,
X2 ..., xn) the feature group (all pixels in pixel matrix) of the image is formed, feature group forms n-dimensional space (Europe
Formula distance is aiming at hyperspace), the condition code (each pixel) in feature group is constituted per one-dimensional numerical value, just
It is that x1 (first pixel) correspondences are one-dimensional, the corresponding two dimensions of x2 (second pixel point) ..., xn (nth pixel point) correspond to n
Dimension.Under n-dimensional space, two image arrays respectively form a point, then calculate this using Euclidean distance formula mathematically
The distance between two points are exactly most matched image apart from reckling.
Euclidean distance formula：
Point A=(x1, x2 ..., xn)
Point B=(y1, y2 ..., yn)
AB^2=(x1-y1) ^2+ (x2-y2) ^2+...+ (xn-yn) ^2
AB is exactly required A, the distance between the point in two hyperspace of B.
Perceptual hash algorithm is as follows：
The first step, minification.
The fastest removal high frequency and details, the method for only retaining structure light and shade is exactly to reduce the size.
Picture is narrowed down to the size of 8x8, in total 64 pixels.Abandon different sizes, the picture difference that proportional band comes.
Second step simplifies color.
By the picture after diminution, switch to 64 grades of gray scales.That is, all pixels point only has 64 kinds of colors in total.
Third walks, and calculates DCT (discrete cosine transform).
DCT is that picture is decomposed frequency aggregation and scalariform shape, although JPEG is used herein using the dct transform of 8*8
The dct transform of 32*32.
4th step reduces DCT.
Although DCT's the result is that 32*32 sizes matrix, as long as we retain the upper left corner 8*8 matrix, this part
Present the low-limit frequency in picture.
5th step calculates average value.
Calculate the average value of all 64 values.
6th step, further decreases DCT.
This is a most important step, and according to the DCT matrixes of 8*8,64 hash values of setting 0 or 1 are more than or equal to DCT
Mean value is set as " 1 ", less than being set as " 0 " for DCT mean values.As a result the low frequency that can not teach that authenticity, can only be rough
Ground teaches that the relative scale relative to average value frequency.As long as the overall structure of picture remains unchanged, hash end values
It is constant.Gamma correction or color histogram can be avoided to be adjusted the influence brought.
7th step calculates cryptographic Hash.
64bit is arranged to 64 longs, the order of combination is not important, as long as it is same to ensure that all pictures all use
Sample order is just.The DCT of 32*32 is converted into the image of 32*32.
It by the comparison result of previous step, combines, just constitutes one 64 integers, here it is this pictures
Fingerprint.The order of combination is not important, as long as ensureing that all pictures all use same order (for example, from left to right, certainly
Push up downward, big-endian).
After obtaining fingerprint, so that it may to compare different pictures, look at how many position is different in 64.In theory
On, this equates calculating " Hamming distance " (Hamming distance).If different data bit is no more than 5, just explanation
Two pictures are much like；If it is greater than 10, just illustrate that this is two different pictures.
SIFT (Scale-invariant feature transform, Scale invariant features transform) feature point extraction is calculated
Method is a kind of algorithm of detection local feature, and the algorithm is by asking characteristic point in a width figure and its retouching in relation to scale and direction
It states son to obtain feature and carry out Image Feature Point Matching, obtains good result.
Basic route：
1. scale space extremum extracting：Search for the picture position on all scales.It is identified by gaussian derivative function latent
The point of interest for scale and invariable rotary.
2. crucial point location：On the position of each candidate, position and ruler are determined by the fine model of a fitting
Degree.The selection gist of key point is in their degree of stability.
3. direction determines：Gradient direction based on image local distributes to each key point position one or more direction.
All subsequent operations to image data are converted both relative to the direction of key point, scale and position, to offer pair
In the invariance of these transformation.
4. key point describes：In the neighborhood around each key point, the ladder of image local is measured on selected scale
Degree.These gradients are transformed into a kind of expression, this deformation and the illumination variation for indicating to allow bigger local shape.
It is described that the image received is divided into multiple identification regions according to matched image template, specifically include following step
Suddenly：
The definition information of parsing and the image template of images match；According to the definition information of image template, image is positioned
Identification region.
S3 carries out content recognition respectively for each identification region.
The configuration information of identification region is converted to data and inserts monitoring data table, and the monitoring data table is sent by S4
To storage database.
The board state collection method based on image recognition further includes：The identified areas of specified image template.
The board state collection method based on image recognition further includes：
The image template of board display is captured, and frame choosing is carried out to typical image region in the image template, and is defined
Data format and the storage side of mapping relations and the typical image region between the typical image region and monitoring data table
The image template defined is sent to template database and stored by formula.
In embodiment, can frame choosing manually or automatically be carried out to the dynamic area in typical image region.
Can by the dynamic area intelligent identifying system in screen automatically to the dynamic area in typical image region into
Row frame selects.The system includes：Video acquisition device, the video image for acquiring board display；Picture acquiring device is used for
Batch picture is obtained from the video image that video acquisition device acquires, and is sent to dynamic area position detecting module；Dynamically
Regional location detection module extracts the pixel value in each region, according to different pictures for dividing region in advance to every pictures
Pixel value in middle the same area identifies dynamic area, and is detected from picture according to the profile and aberration of different zones
Go out position of the dynamic area in picture, and position data of the dynamic area in picture is sent to dynamic area type identification
Module；Dynamic area type identification module, the data type for identifying dynamic area according to optical character recognition method, and
The data type of the position data of dynamic area and dynamic area is sent to database as template；Database, for storing
Template Information is convenient for subsequent inquiry and calling.
Can by the dynamic area intelligent identification Method in screen automatically to the dynamic area in typical image region into
Row frame selects.This approach includes the following steps：The video image of board display is acquired using video acquisition device；It is obtained using picture
It takes device to obtain batch picture from the video image that video acquisition device acquires, and is sent to dynamic area position detection mould
Block；Region is divided in advance to every pictures using dynamic area position detecting module, the pixel value in each region is extracted, according to not
With the pixel value in the same area in picture, dynamic area is identified, and according to the profiles of different zones and aberration from picture
In detect position of the dynamic area in picture, and position data of the dynamic area in picture is sent to dynamic area class
Type identification module；The data class of dynamic area is identified according to optical character recognition method using dynamic area type identification module
Type, and the data type of the position data of dynamic area and dynamic area is sent to database as template；Utilize database
Template Information is stored, subsequent inquiry and calling are convenient for.Dynamic area position detecting module is enhanced using Binarization methods, edge
Any in algorithm and contour detecting algorithm identifies dynamic area, and detects position of the dynamic area in picture.
In practical applications, it is not overlapped as possible between different typical image regions, typical image region shape is regular as possible.
The mapping relations defined between typical image region and monitoring data table are that the typical image region in image is converted to prison
Measured data table, table 1 are monitoring data table.
Table 1
For template database for storing image template, template database can be structural database or non-structural database,
In template database, different image templates is numbered, to distinguish different images template.Image template can be screen
Sectional drawing or record screen file.
To when typical image region progress frame selects in the image sample, absolute pixel may be used in typical image regional location
Or relative position, typical image region shape can be regular rectangular shape, oval even irregular area.
Template database includes：Output information table and the mapping table for indicating template and the mapping relations of image.
The example of image template is as follows：
In addition, template identification area can also be arranged on image template, more easily to distinguish different image templates.
The type in typical image region includes numerical value, text, icon and chart, the data format in the typical image region and
Storage mode is as shown in table 2.
Table 2
Table 3 is the monitoring data table of the embodiment of the present invention.
Table 3
In table, first is classified as timestamp, and secondary series is classified as board state parameter to the 7th.
For embodiment of the method, for simple description, therefore it is all expressed as a series of combination of actions, but this field
Technical staff should know that the embodiment of the present invention is not limited by the described action sequence, because implementing according to the present invention
Example, certain steps can be performed in other orders or simultaneously.Next, those skilled in the art should also know that, specification
Described in embodiment belong to preferred embodiment, necessary to the involved action not necessarily embodiment of the present invention.
Fig. 3 is that a kind of board state acquisition system based on image recognition of the embodiment of the present invention connects block diagram, such as Fig. 3 institutes
Show, the board state acquisition system provided by the invention based on image recognition, including：
Board display image harvester, the image for capturing board display in real time, and the image is sent
To template matches module；Template matches module, for receiving image in real time, by the image template in image and template database
It is compared, and calls the image template to match with image, divided the image received according to the matched image template
For multiple identification regions；Content recognition scheduler module is called for being directed to each identification region in the progress of content recognition engine respectively
Hold identification, is additionally operable to the monitoring data table being sent to storage database；Content recognition engine is used for the configuration of identification region
Information is converted to data filling monitoring data table, and the monitoring data table is sent to content recognition scheduler module.
The board display image harvester based on image recognition further includes：Template definition is used for guide module
Image template is received, frame choosing is carried out to typical image region in the image template, and defines the typical image region and monitoring number
According to the data format and storage mode of mapping relations and the typical image region between table, the image template defined is sent out
It send to template database and is stored；Wherein, board display image harvester is additionally operable to the image of capture board display
The image template is sent to template definition to guide module by template when board initializes.
Board display image harvester is camera or video frequency collection card.Wherein, video frequency collection card can be adopted for vga
Truck can also capture the image of board display in real time by screenshotss software.
Template definition supports the variable-resolution and variable screen size of image to guide module, at this point, only with opposite position
It sets, rather than absolute pixel position carries out the matching of image and image template.
Template matches module further includes identification daily record, records the picture etc. of identification process and None- identified.
Identification region includes numerical identification region, text identification region, icon-based programming region, Table recognition region and curve
Identification region, content recognition engine include numerical value content identification engine, content of text identification engine, iconic content identification engine,
Table content identifies engine and curve content recognition engine.
Content recognition engine passes through Python, R, JAVA, C etc. for number, word, curve, the identification of the contents such as table
Language is realized, is commonly entered as the configuration information in picture+region, is exported as number, word, the recognizable content such as curve point.
Specifically, numerical value content identification engine passes through various OCR (Optical Character Recognition, optics
Character recognition) technology criterion of identification numerical value, and such as Chinese character date recognition special value is defined by masterplate.Content of text identifies
Engine identifies text by standard OCR technique.Iconic content identifies that engine identifies figure by the similarity mode of color/form
Mark.Due to including numerical value, text and icon in table, so table content identification engine passes through various OCR technique criterion of identification
Numerical value, such as Chinese character date recognition special value is defined by masterplate, is identified text by standard OCR technique, is passed through color/shape
The similarity mode of state identifies icon, and identifies table according to the mapping relations of the row/column and tables of data specified in configuration wizard
In row/column.Curve content recognition engine extraction Curves Recognition region passes through curvilinear coordinate system and reference axis scale judgment curves
Curve type in identification region, according to matched image template to the data format of picture point defined in the curve type and
Picture point in Curves Recognition region is converted to data filling monitoring data table and is sent to storage database by storage mode.
The curve type by curvilinear coordinate system and reference axis scale judgment curves identification region specifically includes：
Using 0 point of curvilinear coordinate system as starting point, to the right upwards for just, record screen pixels point sets screen pixels point to (li+1,
hi+1)；Pass through OCR (Optical Character Recognition, optical character identification) technology coordinate identification axis scale；Note
The picture point for recording curve, sets (x, y) picture point to, and x and y are that scale numerical difference or pixel are poor；By screen pixels point,
Curve type in the picture point judgment curves identification region of reference axis scale and curve.
Picture point in the region by Curves Recognition is converted to data filling monitoring data table, specifically includes：For even
Continuous two frame curve identification regions do difference, if indifference, continue waiting for next frame；It, will current time conduct if variant
The computer of latest data point (i+1) records the time；Identify the screen pixels point coordinates of the latest data point of current curves variation
(li+1, hi+1)；Output data (the xl of latest data point is calculated according to the screen pixels point coordinatesi+1, yhi+1) filling monitoring
Tables of data.If x-axis be time shaft, when the corresponding temporal resolution of x-axis unit pixel point be more than the average computer update cycle,
Computer of being then subject to records the time, and otherwise, two times all retain.Wherein, herein, the resolution ratio of numerical value refers to each
The corresponding numerical scale of pixel.Wherein, when x-axis is time shaft, if the corresponding temporal resolution of x-axis unit pixel point is more than
Computer is averaged the update cycle, then timestamp is subject to the computer record time, for example, the x-axis unit pixel point update cycle is
5s, computer update cycle are 2s, then timestamp is subject to the computer record time, primary every 2s records, that is, 2s, 4s,
6s、8s……；If the corresponding temporal resolution of x-axis unit pixel point is averaged the update cycle less than computer, timestamp includes x
The corresponding temporal resolution of axis unit pixel point and computer are averaged the update cycle, for example, the x-axis unit pixel point update cycle is
3s, computer update cycle are 10s, then timestamp is primary every 3s records and is also recorded once every 10s, that is, 3s, 6s,
9s、10s、12s、15s、18s、20s……。
When image template specifies specific content recognition engine to each identification region, specified according to image template will
Each identification region is passed in different content recognition engines and carries out content recognition；When image template is specified not to each identification region
When specific content recognition engine, the content of corresponding acquiescence is searched in board system registration information according to each identification region
Identify engine.Result information after being identified to content recognition engine is checked and is handled, and by the result information according to image
The information of monitoring data table in template, is monitored the generation of tables of data content, to obtain complete monitoring data table and defeated
Go out.
Content recognition engine can be plug-in content recognition engine, and the interface requirement of the content recognition engine is as follows：Have
Unified entrance function (software convention), at least one pictures/images Data entries in suction parameter；Returned data type pair
As and related prompt message (ratio such as whether Success Flag position, warning illustrate information etc.).
In embodiment, it is possible to specify the refreshing frequency of image, it is not necessary to all be studied and judged per frame, to reduce online content identification
Calculate the time.
In addition, under the premise of business or physical logic meet, the consistent content of the refreshing frequency of image is existed as possible
In one monitoring data table, to improve storage efficiency.
For system embodiments, since it is basically similar to the method embodiment, so fairly simple, the correlation of description
Place illustrates referring to the part of embodiment of the method.
The board state collection method based on image recognition and system of the present invention is by establishing template database come online
Matching will be divided into each identification region per frame board state picture, according to the matched image template of institute per frame board state picture,
And different types of content recognition engine is called to carry out content recognition to be directed to each identification region, to obtain number, word and song
The recognizable contents such as line point, not only shared memory space is small, but also convenient for monitoring, at low cost.
System embodiment described above is only schematical, wherein the unit illustrated as separating component can
It is physically separated with being or may not be, the component shown as unit may or may not be physics list
Member, you can be located at a place, or may be distributed over multiple network units.It can be selected according to the actual needs
In some or all of module achieve the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying creativeness
Labour in the case of, you can to understand and implement.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
It is realized by the mode of software plus required general hardware platform, naturally it is also possible to pass through hardware.Based on this understanding, on
Stating technical solution, substantially the part that contributes to existing technology can be expressed in the form of software products in other words, should
Computer software product can store in a computer-readable storage medium, such as ROM/RAM, magnetic disc, CD, including several fingers
It enables and using so that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation
Method described in certain parts of example or embodiment.
In addition, it will be appreciated by those of skill in the art that although some embodiments in this include institute in other embodiments
Including certain features rather than other feature, but the combination of the feature of different embodiment means to be in the scope of the present invention
Within and form different embodiments.For example, in the following claims, embodiment claimed it is arbitrary it
One mode can use in any combination.
Finally it should be noted that：The above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although
Present invention has been described in detail with reference to the aforementioned embodiments, it will be understood by those of ordinary skill in the art that：It still may be used
With technical scheme described in the above embodiments is modified or equivalent replacement of some of the technical features；
And these modifications or replacements, various embodiments of the present invention technical solution that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (10)
1. a kind of board state collection method based on image recognition, which is characterized in that include the following steps：
The image of board display is captured in real time；
Image and the image template in template database are compared, and call the image template to match with image, according to
The image received is divided into multiple identification regions by the matched image template；
Content recognition is carried out respectively for each identification region；
The configuration information of identification region is converted into data and inserts monitoring data table, and the monitoring data table is sent to storage number
According to library.
2. the board state collection method according to claim 1 based on image recognition, which is characterized in that further include：
The image template of board display is captured, and frame choosing is carried out to typical image region in the image template, and defines the allusion quotation
The data format and storage mode of mapping relations and the typical image region between type image-region and monitoring data table, will
The image template defined is sent to template database and is stored.
3. the board state collection method according to claim 2 based on image recognition, which is characterized in that further include：Refer to
Determine the identified areas of image template.
4. the board state collection method according to claim 3 based on image recognition, which is characterized in that described by image
It is compared with the image template in template database, specifically includes following steps：
With the presence or absence of the image template with images match in judge templet database；
When the image template with images match is not present in template database, the process flow of None- identified is carried out；
When there is the image template with images match in template database, the content part of the identification region of image is deducted, with
Obtain the frame of image；
The frame of image and image template are subjected to similarity evaluation；
When there are be first identified Regional Similarity when identified areas, between the frame and image template of image to comment in image template
Valence, then carry out overall similarity evaluation；
When identified areas is not present in image template, overall similarity is directly carried out between the frame and image template of image and is commented
Valence；
The highest image template of frame entirety similarity of selection and image；
When the overall similarity of the image template is higher than threshold value, the content recognition flow of image is carried out.
5. the board state collection method according to claim 4 based on image recognition, which is characterized in that similarity uses
It is any in Euclidean distance algorithm, perceptual hash algorithm or feature point extraction algorithm.
6. the board state collection method according to claim 5 based on image recognition, which is characterized in that it is described according to
The image received is divided into multiple identification regions by the image template matched, and specifically includes following steps：
The definition information of parsing and the image template of images match；
According to the definition information of image template, the identification region of image is positioned.
7. a kind of board state acquisition system based on image recognition, which is characterized in that including：
Board display image harvester, the image for capturing board display in real time, and the image is sent to mould
Plate matching module；
Template matches module compares image and the image template in template database for receiving image in real time, and
The image template to match with image is called, the image received is divided into multiple cog regions according to the matched image template
Domain；
Content recognition scheduler module is called content recognition engine to carry out content recognition, is also used respectively for being directed to each identification region
In the monitoring data table is sent to storage database；
Content recognition engine inserts monitoring data table for the configuration information of identification region to be converted to data, and should
Monitoring data table is sent to content recognition scheduler module.
8. the board display image harvester according to claim 7 based on image recognition, which is characterized in that also wrap
It includes：
Template definition is to guide module, for receiving image template, carries out frame choosing to typical image region in the image template, and fixed
Data format and the storage of mapping relations and the typical image region between the adopted typical image region and monitoring data table
The image template defined is sent to template database and stored by mode；
Wherein,
Board display image harvester is additionally operable to the image template of capture board display, by the figure when board initializes
As template is sent to template definition to guide module.
9. the board display image harvester according to claim 8 based on image recognition, which is characterized in that board
Display image harvester is camera or video frequency collection card.
10. the board display image harvester according to claim 9 based on image recognition, which is characterized in that know
Other region includes numerical identification region, text identification region, icon-based programming region, Table recognition region and Curves Recognition region,
Content recognition engine includes numerical value content identification engine, content of text identification engine, iconic content identification engine, table content knowledge
Other engine and curve content recognition engine.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810124997.XA CN108427959A (en) | 2018-02-07 | 2018-02-07 | Board state collection method based on image recognition and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810124997.XA CN108427959A (en) | 2018-02-07 | 2018-02-07 | Board state collection method based on image recognition and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108427959A true CN108427959A (en) | 2018-08-21 |
Family
ID=63156513
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810124997.XA Pending CN108427959A (en) | 2018-02-07 | 2018-02-07 | Board state collection method based on image recognition and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108427959A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109635265A (en) * | 2018-11-29 | 2019-04-16 | 济南荣耀合创电力科技有限公司 | A kind of test report generation system based on image recognition |
| CN110838191A (en) * | 2019-09-30 | 2020-02-25 | 中车永济电机有限公司 | Novel method for downloading event and fault data of diesel locomotive |
| CN110956168A (en) * | 2019-12-10 | 2020-04-03 | 上海精密计量测试研究所 | Online data acquisition method of single-edition environment test equipment based on image recognition |
| WO2020134710A1 (en) * | 2018-12-28 | 2020-07-02 | 中兴通讯股份有限公司 | Generic data acquisition method and apparatus |
| CN112329851A (en) * | 2020-11-05 | 2021-02-05 | 腾讯科技（深圳）有限公司 | Icon detection method and device and computer readable storage medium |
| CN112348126A (en) * | 2021-01-06 | 2021-02-09 | 北京沃东天骏信息技术有限公司 | Method and device for identifying target object in printed article |
| CN113841156A (en) * | 2019-05-27 | 2021-12-24 | 西门子股份公司 | Control method and device based on image recognition |
| CN115249362A (en) * | 2022-09-20 | 2022-10-28 | 京华信息科技股份有限公司 | OCR table recognition method and system based on connectivity of pixels in stable direction |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104537339A (en) * | 2014-12-19 | 2015-04-22 | 汉王科技股份有限公司 | Information identification method and information identification system |
| CN105005793A (en) * | 2015-07-15 | 2015-10-28 | 广州敦和信息技术有限公司 | Method and device for automatically identifying and recording invoice character strip |
| CN106444648A (en) * | 2016-06-27 | 2017-02-22 | 惠州市物联微电子有限公司 | Production equipment data acquisition apparatus and method for smart factory |
| CN106909941A (en) * | 2017-02-27 | 2017-06-30 | 广东工业大学 | Multilist character recognition system and method based on machine vision |
| CN107545240A (en) * | 2017-07-07 | 2018-01-05 | 杰为软件系统(深圳)有限公司 | A kind of industrial control equipment display screen output picture acquisition system and method |
-
2018
- 2018-02-07 CN CN201810124997.XA patent/CN108427959A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104537339A (en) * | 2014-12-19 | 2015-04-22 | 汉王科技股份有限公司 | Information identification method and information identification system |
| CN105005793A (en) * | 2015-07-15 | 2015-10-28 | 广州敦和信息技术有限公司 | Method and device for automatically identifying and recording invoice character strip |
| CN106444648A (en) * | 2016-06-27 | 2017-02-22 | 惠州市物联微电子有限公司 | Production equipment data acquisition apparatus and method for smart factory |
| CN106909941A (en) * | 2017-02-27 | 2017-06-30 | 广东工业大学 | Multilist character recognition system and method based on machine vision |
| CN107545240A (en) * | 2017-07-07 | 2018-01-05 | 杰为软件系统(深圳)有限公司 | A kind of industrial control equipment display screen output picture acquisition system and method |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109635265A (en) * | 2018-11-29 | 2019-04-16 | 济南荣耀合创电力科技有限公司 | A kind of test report generation system based on image recognition |
| WO2020134710A1 (en) * | 2018-12-28 | 2020-07-02 | 中兴通讯股份有限公司 | Generic data acquisition method and apparatus |
| CN111382731A (en) * | 2018-12-28 | 2020-07-07 | 中兴通讯股份有限公司 | General data acquisition method and device |
| CN113841156A (en) * | 2019-05-27 | 2021-12-24 | 西门子股份公司 | Control method and device based on image recognition |
| CN110838191A (en) * | 2019-09-30 | 2020-02-25 | 中车永济电机有限公司 | Novel method for downloading event and fault data of diesel locomotive |
| CN110956168A (en) * | 2019-12-10 | 2020-04-03 | 上海精密计量测试研究所 | Online data acquisition method of single-edition environment test equipment based on image recognition |
| CN112329851A (en) * | 2020-11-05 | 2021-02-05 | 腾讯科技（深圳）有限公司 | Icon detection method and device and computer readable storage medium |
| CN112348126A (en) * | 2021-01-06 | 2021-02-09 | 北京沃东天骏信息技术有限公司 | Method and device for identifying target object in printed article |
| CN112348126B (en) * | 2021-01-06 | 2021-11-02 | 北京沃东天骏信息技术有限公司 | Method and device for identifying target object in printed article |
| CN115249362A (en) * | 2022-09-20 | 2022-10-28 | 京华信息科技股份有限公司 | OCR table recognition method and system based on connectivity of pixels in stable direction |
| CN115249362B (en) * | 2022-09-20 | 2022-12-27 | 京华信息科技股份有限公司 | OCR table recognition method and system based on connectivity of pixels in stable direction |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108427959A (en) | Board state collection method based on image recognition and system | |
| US9665790B2 (en) | Robust and efficient image identification | |
| US9811749B2 (en) | Detecting a label from an image | |
| WO2019169772A1 (en) | Picture processing method, electronic apparatus, and storage medium | |
| US20130129152A1 (en) | Methods and systems for improving yield in wanted vehicle searches | |
| KR100884066B1 (en) | System and method for comparing image based on singular value decomposition | |
| WO2010114478A1 (en) | Apparatus and methods for analysing goods cartons | |
| CN111639629B (en) | Pig weight measurement method and device based on image processing and storage medium | |
| CN105139011B (en) | A kind of vehicle identification method and device based on mark object image | |
| CN108921160B (en) | Book identification method, electronic equipment and storage medium | |
| CN108287833A (en) | It is a kind of for the art work identification to scheme to search drawing method | |
| CN108446717A (en) | A kind of board state collection method and system based on image recognition | |
| CN104915664B (en) | Contact object identifier obtaining method and device | |
| CN111625687B (en) | Method and system for quickly searching people in media asset video library through human faces | |
| CN111858977B (en) | Bill information acquisition method, device, computer equipment and storage medium | |
| CN112686258A (en) | Physical examination report information structuring method and device, readable storage medium and terminal | |
| CN107203638B (en) | Monitoring video processing method, device and system | |
| CN108921016B (en) | Book score obtaining method based on image recognition, electronic equipment and storage medium | |
| CN110443306B (en) | Authenticity identification method for wine cork | |
| CN117115569A (en) | Automatic object image identification and classification method and system based on machine learning | |
| CN109359616A (en) | A kind of pseudo- splicing small size algorithm for recognizing fingerprint based on SIFT | |
| Chaudhury et al. | Google newspaper search–image processing and analysis pipeline | |
| CN110619056A (en) | Invoice input method, device, equipment and computer storage medium | |
| CN112507999B (en) | Non-invasive user interface input item identification method based on visual characteristics | |
| CN203596856U (en) | Cigarette 32-bit code image acquisition processing device |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180821 |