CN1240021C - Bill image processing equipment - Google Patents
Bill image processing equipment Download PDFInfo
- Publication number
- CN1240021C CN1240021C CN 03148249 CN03148249A CN1240021C CN 1240021 C CN1240021 C CN 1240021C CN 03148249 CN03148249 CN 03148249 CN 03148249 A CN03148249 A CN 03148249A CN 1240021 C CN1240021 C CN 1240021C
- Authority
- CN
- China
- Prior art keywords
- image
- data
- unit
- input
- bill
- 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.)
- Expired - Fee Related
Links
Images
Abstract
The present invention discloses a bill image processing device for an optical character recognition system, which comprises an internal bus, memory units, a core processor and an external bus interface expanding unit, wherein the memory units are mutually connected by the internal bus and are used for storing optical recognition characters (ORC) character template base, a layout template base and a bill image processing program. The external bus interface expanding unit is used for connecting the core processor and a host computer bus in a bridging mode. The core processor is particularly used for processing or recognizing bill images. By adopting the device, the treatment of image data of an OCR system no longer depends on a host computer of the OCR system. Thus, the system has multiple resources for processing the bill images. Thereby, the efficiency of system image treatment can be greatly enhanced.
Description
Technical field
The present invention relates to a kind of image-processing system, especially a kind of device that the bill image is handled of in optical character identification (OCR) system, being used for.
Background technology
In the informationized society that arrives gradually, along with the development of computer networking technology, bigger variation has taken place in the communication media of information, makes the expression of information, propagation also must enter electronic development model.Because the business development of all trades and professions, rolling up of the kind of bill, quantity, handle and preserve the process of these bills and the also more and more miscellaneous and poor efficiency of process of information extraction from these bills, like this, adopt computing machine and optical image information fetch equipment to make up the OCR system, handle and preserve described bill and information wherein becomes inevitable trend in the electronics mode.
Existing OCR system, mainly be by computing machine and optical image fetch equipment and control described computing machine and the optical image fetch equipment constitutes by the control program of required mode work, more particularly, CPU by described control program control computer reads the bill image that needs preservation and handle according to mode designated control optical image fetch equipment, these images are handled according to mode designated by CPU again, obtained the information of respective classes.Because the CPU of OCR system computer not only will be used to handle bill, also to bear the operations such as Network Transmission of management, scheduling and the information of whole OCR system, as management to operating system platform, this makes limited cpu resource only have sub-fraction can be used to handle bill, greatly reduce the treatment effeciency of bill, strengthened the bill processing cost.In the reality, improve the processing power of bill, must will increase computer equipment, thereby cost of investment, maintenance cost, resource consumption of making whole OCR system etc. increase considerably.
In addition, existing security, the stability that comprises the OCR system of computer equipment is vulnerable to the external world or illegally uses and the influence of improper use.Therefore, how to improve the processing power of system, and the cost of investment, maintenance cost, the resource consumption that less increase system are the problems that presses for solution in the practice.The straightforward procedure that addresses the above problem is exactly that the device that is specifically designed to disposal system bill image is set, it is the association's transaction card that is exclusively used in disposal system bill image in the system, yet, people's understanding has been confined to based on programme controlled computer system for a long time, makes the device of processing OCR system bill image of this special use not occur.
Summary of the invention
The objective of the invention is to, a kind of bill image-processing system that is used for Optical Character Recognition system is provided, use this device greatly to improve the bill treatment effeciency of OCR system with less cost.
For achieving the above object, bill image-processing system provided by the invention is used for Optical Character Recognition system, and described Optical Character Recognition system comprises the optical image fetch equipment and comprise the principal computer of bus that described bill image-processing system comprises:
Internal bus; And by the interconnective storage unit that is used for storing optical identification character (OCR) literal template base, space of a whole page template base and bill image processing program of described internal bus; With,
Core processor, be used to receive the layout image data of optical image fetch equipment by the bus input of described principal computer, these space of a whole page layout image data are carried out space of a whole page template base in the space of a whole page pre-stored unit and the template data in the OCR literal template base compares processing, with comparative result or directly pretreated space of a whole page data are fed back to principal computer.
Described storage unit comprises:
The template data storage unit is used to store the data of OCR literal template base and space of a whole page template base;
The template index unit, the numbering and the address that are used to store literal template and image templates provide the index information for the template data inquiry.
Described core processor comprises:
The image pretreatment unit is used for the correcting process before the capable identification of the space of a whole page of input, and extracts the space of a whole page characteristic output of layout image data, and described space of a whole page characteristic comprises image feature data and character features data;
The information comparison recognition unit is used for literal template data and the image templates data of the space of a whole page characteristic of input imagery with storage are in advance compared, and obtains the layout information output of input imagery;
Input and output control unit is used for the layout image data that the optical image fetch equipment is imported by the bus of described principal computer are input to described visual pretreatment unit or the identification of described information comparison
The buffering of unit and sequential control, and the buffering and the sequential control that the layout information data that the characteristic or the described information comparison recognition unit of described visual pretreatment unit output are exported are outputed to principal computer.
Described visual pretreatment unit comprises:
The image dsitortion rectification module, the layout image data that are used for right input are deceived urgent processing and the inclination correction process removed;
Image enhancement and noise cleaning module are used for the layout image data of input are carried out Filtering Processing, remove the ground unrest in the image, and strengthen image;
The image binaryzation processing module, being used for colour or gray scale layout image data conversion through image enhancement and noise cleaning is to have only layout image data black, white two-value;
Word message extracts repairs module, be used for taking assigned address and size from the layout image data of handling through image binaryzation, and comprise the image board face of specifying Word message, and removing takes the noise on the image board face, and repairs the character and picture defective that causes because of noise cleaning;
Module collected in literal, line information, are used for search image characteristic and character features data from binary image plate face or the binary image plate face that takes.
Described information identificating unit comprises:
The image comparing unit is used for the image feature data of space of a whole page characteristic and the template data in the space of a whole page template base are compared, and exports this comparative result under the control of input and output control unit;
Word recognition unit is used for the character features data of space of a whole page characteristic and the template data in the OCR literal template base are compared, and exports this comparative result under the control of input and output control unit.
Described input and output control unit comprises:
Input queue, be used for according to the layout image data of the sequential storage optical image fetch equipment of importing, and export described layout image data to described visual pretreatment unit or described information comparison recognition unit according to the order of storage by the bus input of described principal computer;
Output queue is used for according to the sequential storage image pretreatment unit of output or the visual preprocessed data or the visual layout information data of information comparison recognition unit output, and exports these data to principal computer according to the order of storage;
The I/O management unit is used to control the order input pending layout image data of described input queue according to first in first out, controls described output queue according to the order output pre-service result data of first in first out or the layout information data of image.
Described I/O management unit also comprises:
First determining device is used for finishing the judgement that the layout image data of input queue are exported to visual pretreatment unit or information comparison recognition unit;
Second determining device is used to finish the judgement that the visual preprocessed data of visual pretreatment unit output is exported to output queue or information comparison recognition unit.
Bill image-processing system of the present invention also comprises:
Power Management Unit is used for storage unit and core processor are carried out overcurrent protection, overvoltage protection, anti-interference and antistatic protection, and voltage-stabilizing protection is provided.
Ciphering unit, the layout image data that are used for principal computer is input to described visual pretreatment unit or described information comparison recognition unit are decrypted operation, and visual preprocessed data or visual layout information data that visual pretreatment unit or information comparison recognition unit output to principal computer are carried out cryptographic operation.
Another bill image-processing system that is used for Optical Character Recognition system provided by the invention is the multi-core processor device, and it comprises:
Internal bus; And,
By the interconnective storage unit that is used for storing optical identification character (OCR) literal template base, space of a whole page template base and bill image processing program of described internal bus; With,
At least two core processors, each core processor is used to receive the layout image data of optical image fetch equipment by the bus input of described principal computer, these layout image data are carried out space of a whole page pre-service, compare with space of a whole page template base in the described storage unit and the template data in the OCR literal template base, with comparative result or directly pretreated space of a whole page data are fed back to principal computer;
The external bus interface expanding element is used for the bus of described core processor of bridge joint and principal computer.The type of the bus of described external bus interface expanding element type and principal computer adapts.
Because the present invention adopts the bill image-processing system that comprises core processor to be exclusively used in the bill image of handling Optical Character Recognition system, make the processing of the pictorial data of system no longer rely on the principal computer of system, can make system have more resources like this and be used to handle the bill image, thereby greatly improve the efficient that system image is handled; Simultaneously, when system image processing load is heavier, can increase the quantity of device core processor of the present invention by the external bus interface expanding element, thereby the processing power of higher pictorial data can be provided with lower cost, and then greatly improve the bill treatment effeciency of whole OCR system; In addition, Power Management Unit and cryptographic processing unit that the present invention adopts also help and improve stability of the present invention and security.
Description of drawings
Fig. 1 is the OCR system construction drawing that comprises the first embodiment of the invention block diagram;
Fig. 2 is the template storage unit block diagram that the described embodiment of Fig. 1 adopts;
Fig. 3 is the core processor block diagram that the described embodiment of Fig. 1 adopts;
Fig. 4 is the visual pretreatment unit block diagram that the described core processor of Fig. 3 adopts;
Fig. 5 is the information comparison recognition unit block diagram that the described core processor of Fig. 3 adopts;
Fig. 6 is the input and output control unit block diagram that the described core processor of Fig. 3 adopts;
Fig. 7 is the I/O management unit block diagram that the described input and output control unit of Fig. 6 adopts
Fig. 8 is the OCR system construction drawing that comprises the second embodiment of the invention block diagram;
Fig. 9 is the Power Management Unit block diagram that the described embodiment of Fig. 1 and Fig. 8 adopts;
Figure 10 is 3 * 3 matrix diagram;
Figure 11 is communicated with the piece synoptic diagram for literal;
Figure 12 looks for the horizontal line process flow diagram.
Embodiment
The process that Optical Character Recognition system is carried out the bill processing mainly comprises following components: at first be the identification of carrying out bill, carry out the extraction of information according to the classification of the bill that identifies then, at last the bill that identifies and the information of extraction are carried out needed management.In said process, the data volume that the bill identifying need be handled is maximum, and the resource that takies is also maximum, and this time that has just caused system's principal computer to carry out the bill processing prolongs greatly.Address this problem, one of way is exactly to adopt in system and is exclusively used in association's treating apparatus of handling the bill image and replaces identification of host computer processes bill and billing information extraction etc. to take the more operation of resource preferably, is finished by principal computer and will take the less operation of resource for the management of bill etc.The present invention just provides a kind of such bill image-processing system that is used for Optical Character Recognition system.
Fig. 1 is the OCR system construction drawing that comprises the first embodiment of the invention block diagram.Shown in Figure 1 is whole Optical Character Recognition system, comprises optical image fetch equipment and the principal computer 3 that comprises bus 2 in the system, because the optical image fetch equipment can be regarded as the part of principal computer 3, so do not draw among the figure.Bill image-processing system 1 among Fig. 1 is device of the present invention, and it comprises internal bus 12 and storage unit 11 that links to each other by it and core processor 13; Described storage unit 11 is mainly used in storage OCR literal template base, space of a whole page template base and bill image processing program.The template data of described OCR literal template base, space of a whole page template base be system under the control of principal computer in advance the bill according to standard generate, comprised the feature that characteristic line, unique point and feature Word message etc. in the bill image can be represented the bill image space of a whole page.If system need expand the processing capacity to newly-increased bill, can generate new add in above-mentioned database corresponding of literal template according to new standard ticket at any time with space of a whole page template data.Described core processor 13 is used to receive the layout image data of optical image fetch equipment by the bus input of described principal computer, and these layout image data are carried out space of a whole page pre-service.According to principal computer given processing instruction when the bill layout image data that read by its bus input optical image fetch equipment, to compare through space of a whole page template base in pretreated layout image data of the space of a whole page and the described storage unit and the template data in the OCR literal template base, feed back to principal computer with comparative result or with pretreated layout image data, perhaps directly the pretreated layout image data of the space of a whole page are fed back to principal computer.Mate by layout image data and predefined space of a whole page template data like this, can finish bill space of a whole page recognition function input.Layout image The data BMP form described here can certainly adopt other form, as the TIF form.
In other embodiments of the invention, difference according to processing bill character, also can only preserve in the described storage unit 11 OCR literal template base or space of a whole page template base one of them, this depends on whether bill identification the time only relies on the identification that OCR literal template data or space of a whole page template data just can be finished bill.
Two selectable assembly of elements also are provided among Fig. 1, have been used to bill image-processing system 1 that safety assurance is provided.One is that 15, one of Power Management Unit are encrypting modules 14.Described Power Management Unit 15 is used for storage unit and core processor are carried out overcurrent protection, overvoltage protection, anti-interference and antistatic protection, and voltage-stabilizing protection is provided.Like this, just can prevent that current overload, voltage overload and static from causing device components from being damaged and the voltage reading and writing data error that causes on the low side, the anti-power supply interference capability of device in the intensifier.Described Power Management Unit is with reference to figure 9, and it comprises Voltage stabilizing module 151, filtration module 152, overcurrent protection module 153, overvoltage protective module 154 and electrostatic prevention module 155.Above-mentioned module adopts present general technology and device to realize, for example to adopt models be that the integrated circuit of PSR05 just can be realized for electrostatic prevention module 155, this integrated circuit is produced by PROTEK DEVICE, comprise low electric capacity, high speed TVS (transient voltage inhibition diode), have the ability that in the very short time, voltage is discharged.
In order to improve the security of data manipulation, input to the pending layout image data of bill image-processing system 1 by the bus 2 of principal computer 3, and feed back to after handling by bill image-processing system 1 principal computer the layout image data can with encrypt mode solve, so just need a ciphering unit and principal computer cooperating, therefore, described ciphering unit 14, the layout image data that just are used for principal computer is input to described visual pretreatment unit or described information comparison recognition unit are decrypted operation, and visual preprocessed data or visual layout information data that visual pretreatment unit or information comparison recognition unit output to principal computer are carried out cryptographic operation.Like this, just solved the data manipulation safety problem.
The template storage unit 11 that the described embodiment of Fig. 1 adopts is with reference to figure 2, and it comprises template data storage unit 111, template index unit 112 and control program storage unit 113.In this example, described template data storage unit 111 is used to store the data of OCR literal template base and space of a whole page template base, 112 of template index unit are used to store the numbering and the address of literal template and image templates, like this, when needs obtain template data, at first by template index unit 112 obtain this template data in template data storage unit 111 numbering and address, block size etc. for the index information of template data inquiry, in template data storage unit 111, extract the template corresponding data according to above-mentioned information again.Control program storage unit 113 is mainly used in the storage corresponding control programs.When the present embodiment specific implementation, described template storage unit 11 adopts the storer of two kinds of character, a kind ofly can forever preserve data its content of can making amendment at any time again, as EPROM, EEPROM, FLASH ROM etc., to make up above-mentioned control program storage unit 113; Another kind is a random access memory, as SDRAM, be used to make up template data storage unit 111, template index unit 112, and under the control of control program, provide the space of storing and handling for core processor 13 by the layout image data of described principal computer bus input.
The core processor that the described embodiment of Fig. 1 adopts is with reference to figure 3.The described core processor 13 of Fig. 3 comprises that the inside of visual pretreatment unit 131, information comparison recognition unit 132, input and output control unit 133 and core processor 13 is used for the bus 134 of exchanges data.Described visual pretreatment unit 131 is used for the layout image data of input are discerned preceding correcting process, to eliminate the noise that comprises in the pending layout image data, guarantees accurately to discern bill.The layout image extracting data that this unit also is used for after correcting process goes out the output of space of a whole page characteristic, carries out space of a whole page identification in order to offer system's principal computer or described information comparison recognition unit 132.Space of a whole page characteristic described here comprises image feature data and character features data, certainly, when different bills being discerned processing, different attribute according to bill, described characteristic also can be not limited to above-mentioned two kinds, for example when colored bill is discerned, can adopt gray feature data or the like.
Described information comparison recognition unit 132 is used to obtain the information data of the required identification space of a whole page, promptly provides the recognition result of the space of a whole page of discerning.For this reason, this unit compares the space of a whole page characteristic of described visual pretreatment unit 131 outputs with the literal template data and the image templates data of storing in advance, determine template with described space of a whole page characteristic optimum matching, promptly obtained the layout information data of input imagery, and this information data has been exported to subsequent treatment such as system's principal computer manages.
In the reality, for the identification that makes the bill space of a whole page has higher efficient and dirigibility, system's host computer chance is adjusted the division of labor of data processing in the bill space of a whole page identifying according to the quality of resource utilization ratio or bill image to be identified, for example, when transmitting pending layout image data to the bill image-processing system, send processing instruction simultaneously, the result of the layout image data that indication need obtain, i.e. indication is only carried out the pre-service of visual pretreatment unit 131 to the layout image data that needs are handled, or indicate the layout image data that needs are handled to carry out space of a whole page identification, so just relate to the different pieces of information current control in core processor 13.In addition, also relate to the input of pending layout image data, and the buffer problem the during output of result data.For this reason, in the present embodiment, the input and output control unit 133 that provides, the optical image fetch equipment is input to the storage buffering and the sequential control of described visual pretreatment unit 131 or described information comparison recognition unit 132 by the layout image data of described principal computer bus input, and the storage buffering and the sequential control that the layout information data of the characteristic of described visual pretreatment unit output or the output of described information comparison recognition unit are outputed to principal computer.That is to say, input and output control unit 133 is judged the pending layout image data of input according to the indication of system's principal computer, to determine that these data are to be input to described visual pretreatment unit 131 to carry out pre-service or be directly inputted to after described information comparison recognition unit 132 compares or be input to described visual pretreatment unit 131 and carry out pre-service, are input to described information comparison recognition unit 132 again and compare; Simultaneously, when the output result data, the characteristic that judgement should be exported described visual pretreatment unit output still is the layout information data of described information comparison recognition unit output.
When adopting system shown in Figure 1 to carry out the identification of bill, can adopt following process: at first under the control of principal computer 3, the optical image fetch equipment will scan to bill image to be identified be input to described input and output control unit 133 by its bus 2, judge whether that according to the indication of principal computer needs carry out the pretreatment operation of image (by the quality decision of principal computer according to bill image to be identified) by this unit 133, if do not need, illustrate that bill image to be identified only needs identification, at this moment, described input and output control unit 133 sends to information comparison recognition unit 132 with this bill pictorial data to be identified, finish the comparison or the matching operation of pictorial data by it, obtain the information data of the required identification space of a whole page, then the information data after the above-mentioned identification is as a result of sent to described input and output control unit 133, under the control of this unit 133, send to according to principal computer 3 by bus 2; If judgement through described input and output control unit 133, need carry out the pretreatment operation of image, at this moment, described input and output control unit 133 sends to visual pretreatment unit 131 with this bill pictorial data to be identified, by it this bill pictorial data to be identified is discerned preceding image rectification, noise cleaning, processing such as feature extraction, and then judge whether the identification of comparing of this revised pictorial data, if do not need, directly should send to described input and output control unit 133 by revised image, under the control of this unit 133, send to according to principal computer 3 by bus 2, otherwise, will be under the control of described input and output control unit 133, should revised pictorial data, the characteristic information that promptly extracts sends to information comparison recognition unit 132, after finishing the comparison or matching operation of pictorial data by it, information data after the identification that obtains as a result of by described input and output control unit 133, is sent to according to principal computer 3 through bus 2.Need explanation be, the identification of adopting system shown in Figure 1 to carry out bill in the reality is not limited only to said process, and for example, can judge whether earlier need be to the picture information to be identified identification of comparing.
Specifically, in the present embodiment, described visual pretreatment unit 131 further comprises: image dsitortion rectification module 1311, image enhancement and noise cleaning module 1312, image binaryzation processing module 1313, Word message extract repairing module 1314 and literal, line information are collected module 1315, with reference to figure 4.
Described image dsitortion rectification module 1311, the layout image data that are used for right input are carried out black surround excision processing and inclination correction process.In this example, the black surround excision is that the binary picture after handling at image binaryzation processing module 1313 is carried out, specific practice is, space of a whole page level is divided into two zones up and down, in each zone respectively from the left side, the right begins to handle, whole like this zone can be divided into upper left district, lower-left district, upper right district and bottom right district handles, and adopts identical rule to line by line scan respectively in each district.By scanning, to the row that each scanned, according to black row judgment rule determine whether this row is black row, promptly as long as the continuous white point number in this row greater than white noise gap width given in advance, then should capablely not be black row, otherwise went for black.According to the scanning result of each row,, determine black surround and the removal that to remove again according to the judgment rule of black surround.Described image is corrected, and is that the inclination image that scanning produces is corrected.Carrying out image when tilt correcting, can carry out according to sciagraphy, concrete grammar is: one, the angle of inclination of computational picture, two, rotated image.
The method of calculating the angle of inclination is: one, rough search can be 2 degree with step-length to+30 degree scopes at-30 degree usually, searches for; Two, precise search is that 0.1 degree carries out precise search with step-length in the 2 degree scopes that rough search obtains; Mainly be to utilize direction projection statistics stain number when angle searching, promptly, add up in the view picture image stain number of each row in the direction along some angle traversal images, utilize then and add up the stain histogram calculation variance that obtains, the direction of choosing the variance maximum is as the pitch angle.
Described image enhancement and noise cleaning module 1312 are used for the layout image data of input are carried out Filtering Processing, removed the ground unrest of image, and strengthen image.Filtering and noise reduction sound described here is meant and removes unnecessary lines, striped in the image, and carries out smoothing processing, the processing of desalination filter, the processing of enriching filter, the cleaning smoothing processing of character and mend broken string processing etc.The described lines that go, be on image, be the bitmap image, the judgement of lining by line scan, the wide lines of single pixel that search exists, again the wide lines of all single pixels are merged into the thick lines with actual pixels width by adjacent rule, judge then whether these thick lines satisfy the removal condition, if satisfy, explanation is the unnecessary lines in the pattern recognition process, it is removed all lines that satisfy condition up to elimination.If in going the lines process, the part effective information on the character is removed, for guaranteeing the complete of character, at this moment, also to carry out repairing to impaired character.Concrete mending course is that near the distribution situation of the character of lines has been removed in scanning, detects the particular location that needs to repair character, repairs according to the average length of lines length about the damaged part then.
The described striped that goes is with by the scanning to image, length in the delegation is not more than the lines removal of given width.
The smoothing processing of described character is meant for the point in the character to be divided into a little and to mend some two rules that visual bitmap pointwise is judged rule induction is 3 * 3 matrixes, with reference to Figure 10, the central point of this matrix is a judging point.The coordinate of supposing this point for (I, J), (I J) is white point, if 8 some stains are many on every side for they, such as (I-1, J-1), (I, J-1), (I+1, J-1), (I-1, J), (I+1 J) is stain, and then this point should be mended and be stain; Otherwise (I J) is stain, if 8 some white points are many around it, thinks that then this point is a noise point, should remove.
Described desalination filter is handled, and is to adopt corroding method to eliminate object boundary point in the image.If structural element is got 3 * 3 stain piece, corrosion will make the border of object reduce by a pixel along periphery.If between two objects tiny connection is arranged, when structural element is enough big, two objects can be separated so by erosion operation.The expression formula of erosion algorithm is: X-S=∩ X[s] |-s ∈ S}, X are target image, and S is a structural element.At first read in the some pixel value of (being called current point) in the former image, getting with this point is 3 * 3 matrixes at center, if this is a stain, and 8 points not all are stains on every side, and then this point is composed into white point, promptly erodes.
Described enriching filter is handled, and is used to strengthen the brightness of image, improves the contrast of image.Concrete method equation expression is: X+S=∪ X[s] | s ∈ S}, X are target image, and S is a structural element.At first read in the some pixel value of (being called current point) in the former image, getting with this point is 3 * 3 matrixes at center, if this is a white point, and 8 points not all are white points on every side, and then this point is composed into stain, promptly expands.Also can adopt the disposal route of light enriching in the reality, i.e. the shape of limiting structure element S.Such as the expansion process left that makes progress, at first read in the some pixel value of (being called current point) in the former image, get with this point is 3 * 3 matrixes at center, if this is a white point, and left side point, upper left point and last edge point not all are white point (promptly only judging 3 points), then this point is composed into stain, promptly expands.
Cleaning is level and smooth to be handled with mending to break, and its rule also reduces 3 * 3 matrixes, and the central point of this matrix is a judging point.Smoothing processing is considered is a little, and promptly stain bleaches a little, and what mend that broken string considers is to mend point, i.e. white point blackening point.Mend the broken string processing rule, the number of stains in the middle of 8 points around the judgment matrix central point at first, next judges each stain position, whether this point is become stain according to the number and the determining positions of above-mentioned stain; Smoothing processing is at first considered on every side white point number in 8 points, considers each white point position relation then, whether this point is become white point according to the number and the determining positions of above-mentioned white point.
Described image binaryzation processing module 1313, being used for colour or gray scale layout image data conversion through image enhancement and noise cleaning is to have only layout image data black, white two-value.Specifically can adopt such method to realize: at first to carry out Gauss's smothing filtering,, prevent to produce isolated white point and stain after the binaryzation to remove white point and stain noise; Secondly determine the prospect of view picture image and the gray-scale value of background, calculate the binaryzation threshold values, described threshold values can adopt two kinds, and the one, overall static threshold values, the 2nd, local dynamic thresholding; Reset at last the gray-scale value of view picture image, gray-scale value is made as white point greater than the picture element of threshold values, otherwise is made as stain according to the size of threshold values.
Described Word message extracts repairs module 1314, be used for taking assigned address and size from the layout image data of handling through image binaryzation, and comprise the image board face of specifying Word message, and remove noises such as the lines on the image board face of taking, blackspot, and repair the character and picture defective that causes because of noise cleaning, and to the literal rectification etc. of tilting.Literal is after binaryzation in the image, therefore the track of literal is made up of the stain of a series of adjacent arrangements certainly, can classify adjacent stain as one group, and each stain of organizing adjacent arrangement can surround it with a rectangle, form the connection piece, with reference to Figure 11.The method that Word message extracts is as follows:
1, determines the approximate range of required literal;
2, take the image of specified scope;
3, remove lines and shading;
4, search all connection pieces;
5, some arrangements of merging with good conditionsi are communicated with piece closely;
6, classify the connection piece of specified altitude assignment and length as a set A;
7, the Word message of in set A, searching.
Because some literal intersect with lines and shading, in the superincumbent operation, because the step of lines and shading in the existence removal literal, so when removing lines and shading, the part of literal can be erased, in order to increase the accuracy of pattern recognition, also need to carry out literal and repair, concrete method for repairing and mending is with reference to top described.
Module 1315 collected in described literal, line information, are used for search image characteristic and character features data from binary image plate face or the binary image plate face that takes.In this example, this module 1315 is used for searching for the literal of designated shape, size on the image after the binaryzation, and search specifies the lines at angle of inclination, for example horizontal line and vertical line, with the Word message that searches and line information as character features data and image feature data.
The extraction of Word message can be extracted from the connection set of blocks that described Word message extraction repairing module 1314 provides; By the rotation of the space of a whole page, can realize the mutual conversion of horizontal line and vertical line, therefore searching of horizontal line and vertical line can be summed up as and search horizontal line or search vertical line.To search horizontal line is example, with reference to Figure 12.
Before beginning to look for the line operation, lines set variable and current lines variable should be set usually, so that the lines that find are preserved.At first treat the test space of a whole page and carry out point by point scanning line by line, when scanning stain b_dot, change steps A 17 and carry out the stain processing, when scanning white point w_dot, change steps A 13 and carry out the white point processing in steps A 11.If that scan is stain b_dot, whether be that (whether it is stain that left end point is characterized as this point to a line segment left end point, and the front a bit is a white point, can be a line segment left end point by above-mentioned feature judging point b_dot therefore at steps A 17 judging point b_dot.), if, begin to look for the line operation in steps A 18, promptly begin this point is charged to current lines variable, change the operation that steps A 11 continues the scanning subsequent point then; Otherwise, whether at steps A 19 judging point b_dot is a bit (entering when looking for the line operation in the line segment, if this point is a stain, can determine that it is a bit in the line segment), if, directly this point is charged to current lines variable, change steps A 11 then and continue the next point of scanning in steps A 20; Otherwise, whether be the end of delegation at steps A 21 judging point b_dot, and be not the point on the line, if, illustrate that this point may be a noise spot, irrelevant with the lines that will search, at this moment then change steps A 11 and carry out next line scanning; Otherwise, illustrate whether some b_dot are the end of delegation, and be the point on a line, therefore change steps A 22 and do and look for the line processing that finishes.
If that scan is white point w_dot, whether at steps A 13 judging point w_dot is the white point of a line segment interrupt line part, if, the processing of need breaking, therefore in steps A 15 processing of breaking, broken string in this example handle be with a w_dot as stain b_dot, change steps A 12 then and carry out stain and handle; Otherwise whether found line before steps A 14 judging point w_dot, promptly whether the point before this point is the ending of a line, if not, illustrates that this point is common white point, irrelevant with lines, changes steps A 11 this moment and continues the follow-up point of scanning; Otherwise, whether be the end point of a line segment at steps A 16 judging point w_dot, promptly the point before this point is the point of a line, if change steps A 22 and do and look for the line processing that finishes.
Handle at 22 pairs of lines that find of steps A, judge the whether end of scan of the space of a whole page to be tested in steps A 23 then,, change step 11 and continue scanning, otherwise finish scanning, carry out the subsequent treatment of the space of a whole page to be identified if finish.
In this example, whether according to following step judging point w_dot is the white point of a line segment interrupt line part: in the scope that broken string length allows, with the current behavior datum line, be offset nominated bank up and down, begin to search stain from the horizontal ordinate of this white point, if in this scope, do not find stain, judge that then this point is the end of a line segment, otherwise judge that this white point is a broken string part in the line.
The lines of steps A 22 are handled, and are line segment to be carried out the judgement of qualification, if what find is qualified lines, it is saved in the lines set variable, the further feature information that also comprises lines of Bao Cuning simultaneously as parameters such as the length of lines, coordinates, otherwise is abandoned this underproof lines.In this example to the judgement of lines qualification by with relatively the finishing of empirical value, for example, preestablish the length empirical value of hyphen lines, as 9 bits, and the length empirical value of long ledgement, as 120 bits.In the vertical line of down-stream is searched, also to as 70 bits, compare, thereby finish the judgement of vertical line qualification according to the length empirical value of predefined vertical moulding.The main effect of steps A 22 is noise lines of removing in letter in the layout image or the Chinese character, is convenient to follow-up identification.
Need to prove, visual pretreatment unit 131 included image dsitortion rectification module 1311, image enhancement and noise cleaning module 1312, image binaryzation processing module 1313, Word messages shown in Fig. 4 extract repairing module 1314 and literal, line information are collected module 1315, and different choice is used according to the actual requirements.For example, the bill of Chu Liing is a standard black white if desired, can not adopt image binaryzation processing module 1313; If the nominal value image that system scanned can guarantee image sensor and distort in the error range that sets in advance, also can not adopt image dsitortion rectification module 1311, or the like.
The structure of described information identificating unit 132 is with reference to figure 5, and it comprises: visual comparing unit 1321 and word recognition unit 1322.When the layout image data of visual pretreatment unit 131 outputs are carried out matching treatment, can be according to the characteristic of its output, select visual comparing unit 1321 or word recognition unit 1322 or select the two to carry out matching operation simultaneously, can handle a greater variety of bills like this.Specifically, described visual comparing unit 1321 is used for the image feature data of space of a whole page characteristic and the template data in the space of a whole page template base are compared, and exports this comparative result under the control of input and output control unit; Described word recognition unit 1322 is used for the character features data of space of a whole page characteristic and the template data in the OCR literal template base are compared, and exports this comparative result under the control of input and output control unit.In the present embodiment, above-mentioned comparative result at first is sent to described input and output control unit, according to the output order, sends system's principal computer by this unit to.
When the literal that it may be noted that in to layout image to be comprised was handled, if described literal is a graphic form, then the identification of this literal was handled and is finished by visual comparing unit 1321; If described literal is the OCR form, then the identification of this literal is handled and is finished by word recognition unit 1322.
When adopting described information identificating unit 132 to carry out the comparison identifying operation of graph data, at first the characteristic according to visual pretreatment unit input judges whether that needs carry out the comparison identifying operation of pictorial data, if desired, promptly, comprise the image feature data in the characteristic, therefore carry out the matching operation of the template data in image feature data and the space of a whole page template base by visual comparing unit 1321, after obtaining the result, the characteristic according to visual pretreatment unit input judges whether that needs carry out literal identification again; If do not need to carry out the comparison identifying operation of pictorial data, then the characteristic of directly importing according to visual pretreatment unit judges whether that needs carry out literal identification, literal identifying operation if desired, then by described word recognition unit 1322 carry out the character features data with OCR literal template base in the matching operation of template data.
The structure of described input and output control unit 133 is with reference to figure 6, and it comprises: input queue 1333, output queue 1332 and I/O management unit 1331.Described input, output queue are set, and purpose is to realize realizing storage, transmitting control, so that device of the present invention has stronger bill image processing ability for the data of input, output.Described input queue 1333, the bill pictorial data to be identified that is used for the buffer memory input, export described layout image data to described visual pretreatment unit or described information comparison recognition unit promptly according to the layout image data of the sequential storage optical image fetch equipment of importing, and according to the order of storage by described principal computer bus input; Described output queue 1332, be used for buffer memory and export to the pre-service result or the recognition result data of system's principal computer, export these data to principal computer promptly according to the sequential storage image pretreatment unit of output or the visual preprocessed data or the visual layout information data of information comparison recognition unit output, and according to the order of storage;
Above-mentioned input, output queue preferred design become fifo queue, also do not get rid of certainly and use other formation also can reach same effect, as circle queue.
Described I/O management unit 1331, be used to control the order input pending layout image data of described input queue according to first in first out, and control described output queue according to the order output pre-service result data of first in first out or the layout information data of image, its structural drawing is with reference to figure 7.For realizing these control operations, in the present embodiment, described I/O management unit 1331 comprises following several sections: controller 13311 is used for realizing above-mentioned input, output data storage, displacement and the output function in respective queue; Because in core processor, the processing of data and the flow direction are unfixed, therefore, first determining device 13312 also is provided in described I/O management unit 1331, has been used for finishing the layout image data of input queue are exported to visual pretreatment unit or information comparison recognition unit with described data when output judgement; Second determining device 13313 is used to finish the judgement that the visual preprocessed data of visual pretreatment unit output is exported to output queue or information comparison recognition unit.
In other embodiments of the invention, different according to the data processing in the core processor and the flow direction, described I/O management unit 1331 does not comprise described first determining device 13312 and described second determining device 13313, or includes only one of them.
Fig. 8 is the OCR system construction drawing that comprises the second embodiment of the invention block diagram.Comprise optical image fetch equipment and the principal computer 3 that comprises bus 2 in the system shown in Figure 8, and bill image-processing system 1 of the present invention.In described bill image-processing system 1, comprise internal bus 12 and the storage unit 11 that links to each other by it and 4 core processors 13 and external bus interface expanding element 4.Compare with embodiment 1 shown in Figure 1, increased external bus interface expanding element 4 in this example, be used for described core processor of bridge joint and principal computer bus, to expand more core processor 13.When the core processor 13 that is adopted is 2 when above, just need to adopt external bus interface expanding element 4.Therefore, in other embodiment of the present invention that adopts the external bus interface expanding element, the core processor 13 that is connected is 2, also can be 3,5,6 etc., and this depends on pictorial data processing power required for the present invention.Described storage unit 11 is mainly used in storage OCR literal template base, space of a whole page template base and bill image processing program.The template data of described OCR literal template base, space of a whole page template base be system under the control of principal computer in advance the bill according to standard generate, comprised the feature that characteristic line, unique point and feature Word message etc. in the bill image can be represented the bill image space of a whole page.In described 4 core processors 13, each core processor all is used to receive the layout image data of optical image fetch equipment by described principal computer bus input, and these layout image data are carried out space of a whole page pre-service.According to principal computer given processing instruction when the bill layout image data that read by its bus input optical image fetch equipment, to compare through space of a whole page template base in pretreated layout image data of the space of a whole page and the described storage unit and the template data in the OCR literal template base, with comparative result or directly pretreated layout image data are fed back to principal computer.Mate by layout image data and predefined space of a whole page template data like this, can finish bill space of a whole page recognition function input.
The type of the type of described external bus interface expanding element 4 and principal computer bus adapts or is complementary.For example adopt PCI, CPCI, SCSI, USB, general external bus interfaces such as 1349.
4 core processors in this example can realize that the pictorial data that walks abreast handles, thereby improve the ability of data processing, the control method that is adopted is, the same set of handling procedure of operation in described 4 core processors, upper host computer chance is assigned bill image processing task to idle relatively core processor processing according to the size of the operating load of each core processor.In the reality, be not limited to this method, for example, can make each core processor finish different processing capacities, line production, the result that previous DSP finishes are given next DSP and are handled, and realize the parallel work-flow of 4 core processors thus.
In addition, also adopted in this example to be used for storage unit and core processor are carried out overcurrent protection, overvoltage protection, anti-interference and antistatic protection, and the Power Management Unit 15 that voltage-stabilizing protection is provided.With being used for principal computer is input to the visual pretreatment unit of each core processor or the layout image data of described information comparison recognition unit are decrypted operation, and the visual preprocessed data that the visual pretreatment unit or the information comparison recognition unit of each core processor outputed to principal computer or the visual layout information data ciphering unit 14 that carries out cryptographic operation.
Need to prove, compare with first embodiment, in second embodiment, increased external bus interface expanding element 4, therefore also expanded the quantity of core processor 13, but, the structure of storage unit 11, core processor 13, Power Management Unit 15 and ciphering unit 14 all is identical, is not therefore giving unnecessary details.
In the described embodiment of the invention described above, the parts of most critical are core processor, can adopt digital signal processor DSP to realize in the reality, for example adopt the TMS320C6415 series DSP of TI company.
Claims (13)
1, a kind of bill image-processing system that is used for Optical Character Recognition system, described Optical Character Recognition system comprise the optical image fetch equipment and comprise the principal computer of bus, it is characterized in that described bill image-processing system comprises:
Internal bus; And,
By the interconnective storage unit that is used for storing optical identification character (OCR) literal template base, space of a whole page template base and bill image processing program of described internal bus; With,
Core processor, be used to receive the layout image data of optical image fetch equipment by the bus input of described principal computer, these space of a whole page layout image data are carried out space of a whole page template base in the space of a whole page pre-stored unit and the template data in the OCR literal template base compares processing, with comparative result or directly pretreated space of a whole page data are fed back to principal computer.
2, bill image-processing system according to claim 1 is characterized in that, described storage unit comprises:
The template data storage unit is used to store the data of OCR literal template base and space of a whole page template base;
The template index unit, the numbering and the address that are used to store literal template and image templates provide the index information for the template data inquiry.
3, bill image-processing system according to claim 1 is characterized in that, described core processor comprises:
The image pretreatment unit is used for the correcting process before the capable identification of the space of a whole page of input, and extracts the space of a whole page characteristic output of layout image data, and described space of a whole page characteristic comprises image feature data and character features data;
The information comparison recognition unit is used for literal template data and the image templates data of the space of a whole page characteristic of input imagery with storage are in advance compared, and obtains the layout information output of input imagery;
Input and output control unit, be used for the layout image data that the optical image fetch equipment is imported by the bus of described principal computer are input to the buffering and the sequential control of described visual pretreatment unit or described information comparison recognition unit, and the buffering and the sequential control that the layout information data that the characteristic or the described information comparison recognition unit of described visual pretreatment unit output are exported are outputed to principal computer.
4, bill image-processing system according to claim 3 is characterized in that, described visual pretreatment unit comprises:
The image dsitortion rectification module, the layout image data that are used for right input are carried out black surround excision processing and inclination correction process;
Image enhancement and noise cleaning module are used for the layout image data of input are carried out Filtering Processing, remove the ground unrest in the image, and strengthen image;
The image binaryzation processing module, being used for colour or gray scale layout image data conversion through image enhancement and noise cleaning is to have only layout image data black, white two-value;
Word message extracts repairs module, be used for taking assigned address and size from the layout image data of handling through image binaryzation, and comprise the image board face of specifying Word message, and removing takes the noise on the image board face, and repairs the character and picture defective that causes because of noise cleaning;
Module collected in literal, line information, are used for search image characteristic and character features data from binary image plate face or the binary image plate face that takes.
5, bill image-processing system according to claim 3 is characterized in that, described information identificating unit comprises:
The image comparing unit is used for the image feature data of space of a whole page characteristic and the template data in the space of a whole page template base are compared, and exports this comparative result under the control of input and output control unit;
Word recognition unit is used for the character features data of space of a whole page characteristic and the template data in the OCR literal template base are compared, and exports this comparative result under the control of input and output control unit.
6, bill image-processing system according to claim 3 is characterized in that, described input and output control unit comprises:
Input queue, be used for according to the layout image data of the sequential storage optical image fetch equipment of importing, and export described layout image data to described visual pretreatment unit or described information comparison recognition unit according to the order of storage by the bus input of described principal computer;
Output queue is used for according to the sequential storage image pretreatment unit of output or the visual preprocessed data or the visual layout information data of information comparison recognition unit output, and exports these data to principal computer according to the order of storage;
The I/O management unit is used to control the order input pending layout image data of described input queue according to first in first out, controls described output queue according to the order output pre-service result data of first in first out or the layout information data of image.
7, bill image-processing system according to claim 6 is characterized in that, described I/O management unit also comprises:
First determining device is used for finishing the judgement that the layout image data of input queue are exported to visual pretreatment unit or information comparison recognition unit;
Second determining device is used to finish the judgement that the visual preprocessed data of visual pretreatment unit output is exported to output queue or information comparison recognition unit.
8, according to claim 1,2,3,4,5,6 or 7 described bill image-processing systems, it is characterized in that described device also comprises:
Power Management Unit is used for storage unit and core processor are carried out overcurrent protection, overvoltage protection, anti-interference and antistatic protection, and voltage-stabilizing protection is provided.
9, bill image-processing system according to claim 8, it is characterized in that, described device also comprises ciphering unit, the layout image data that are used for principal computer is input to described visual pretreatment unit or described information comparison recognition unit are decrypted operation, and visual preprocessed data or visual layout information data that visual pretreatment unit or information comparison recognition unit output to principal computer are carried out cryptographic operation.
10, a kind of bill image-processing system that is used for Optical Character Recognition system, described Optical Character Recognition system comprise the optical image fetch equipment and comprise the principal computer of bus, it is characterized in that described bill image-processing system comprises:
Internal bus; And,
By the interconnective storage unit that is used for storing optical identification character (OCR) literal template base, space of a whole page template base and bill image processing program of described internal bus; With, at least two core processors, each core processor is used to receive the layout image data of optical image fetch equipment by the bus input of described principal computer, these layout image data are carried out space of a whole page pre-service, compare with space of a whole page template base in the described storage unit and the template data in the OCR literal template base, with comparative result or directly pretreated space of a whole page data are fed back to principal computer;
The external bus interface expanding element is used for the bus of described core processor of bridge joint and principal computer.
11, bill image-processing system according to claim 10 is characterized in that, described device also comprises:
Power Management Unit is used for storage unit and core processor are carried out overcurrent protection, overvoltage protection, anti-interference and antistatic protection, and voltage-stabilizing protection is provided.
12, bill image-processing system according to claim 11, it is characterized in that, described device also comprises ciphering unit, be used for principal computer is input to the visual pretreatment unit of each core processor or the layout image data of described information comparison recognition unit are decrypted operation, and visual preprocessed data or visual layout information data that the visual pretreatment unit or the information comparison recognition unit of each core processor outputs to principal computer are carried out cryptographic operation.
13, bill image-processing system according to claim 10 is characterized in that, the type of the bus of described external bus interface expanding element type and principal computer adapts.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 03148249 CN1240021C (en) | 2003-06-27 | 2003-06-27 | Bill image processing equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 03148249 CN1240021C (en) | 2003-06-27 | 2003-06-27 | Bill image processing equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1482572A CN1482572A (en) | 2004-03-17 |
| CN1240021C true CN1240021C (en) | 2006-02-01 |
Family
ID=34156233
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 03148249 Expired - Fee Related CN1240021C (en) | 2003-06-27 | 2003-06-27 | Bill image processing equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1240021C (en) |
Families Citing this family (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102236782A (en) * | 2010-04-30 | 2011-11-09 | 陈斌 | Character identification method for certificates |
| CN102156866A (en) * | 2011-03-09 | 2011-08-17 | 深圳百维达科技有限公司 | Road sign recognition system and method |
| CN102968638B (en) * | 2011-08-31 | 2016-06-08 | 上海夏尔软件有限公司 | Based on the method that the image definition of keyword optical character recognition judges |
| JP2013109690A (en) * | 2011-11-24 | 2013-06-06 | Oki Electric Ind Co Ltd | Business form data input device, and business form data input method |
| CN103034848B (en) * | 2012-12-19 | 2016-07-06 | 方正国际软件有限公司 | A kind of recognition methods of form types |
| CN103093219A (en) * | 2013-01-28 | 2013-05-08 | 中国航空结算有限责任公司 | Ticket number recognition method based on optical character recognition (OCR) technology |
| CN103440507A (en) * | 2013-09-03 | 2013-12-11 | 北京中电普华信息技术有限公司 | Bill information verifying device and method for verifying bill information |
| CN105095842B (en) * | 2014-05-22 | 2018-12-11 | 口碑控股有限公司 | A kind of method and apparatus of the information identification of document |
| CN104537339B (en) * | 2014-12-19 | 2018-04-10 | 汉王科技股份有限公司 | Information identifying method and information identification system |
| CN104966348B (en) * | 2015-06-18 | 2017-10-24 | 深圳辰通智能股份有限公司 | A kind of bill images key element integrality detection method and system |
| CN105069898B (en) * | 2015-07-15 | 2018-03-30 | 广州敦和信息技术有限公司 | The method and device that a kind of invoice brief note is automatically positioned |
| CN105956590A (en) * | 2016-04-27 | 2016-09-21 | 泰合鼎川物联科技（北京）股份有限公司 | Character recognition method and character recognition system |
| CN106022364B (en) * | 2016-05-13 | 2019-05-10 | 邓昌顺 | A kind of bill classification method |
| CN107368800B (en) * | 2017-07-13 | 2020-02-21 | 上海携程商务有限公司 | Order confirmation method, system, equipment and storage medium based on fax identification |
| CN111091499B (en) * | 2018-10-24 | 2023-05-23 | 方正国际软件(北京)有限公司 | Mobile terminal image correction method and device |
| CN109711396A (en) * | 2018-11-12 | 2019-05-03 | 平安科技（深圳）有限公司 | Generation method, device, equipment and the readable storage medium storing program for executing of OCR training sample |
| CN110188755B (en) * | 2019-05-30 | 2021-09-07 | 北京百度网讯科技有限公司 | Image recognition method and device and computer readable storage medium |
-
2003
- 2003-06-27 CN CN 03148249 patent/CN1240021C/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN1482572A (en) | 2004-03-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1240021C (en) | Bill image processing equipment | |
| CN104361312B (en) | A kind of method and apparatus that character recognition is carried out to image | |
| CN107766854B (en) | Method for realizing rapid page number identification based on template matching | |
| CN106960208A (en) | A kind of instrument liquid crystal digital automatic segmentation and the method and system of identification | |
| CN105283884A (en) | Classifying objects in digital images captured using mobile devices | |
| CN1519768A (en) | Method and device for correcting image askew | |
| CN108038484B (en) | Method for quickly identifying hollow verification code | |
| CN108446702B (en) | Image character segmentation method, device, equipment and storage medium | |
| CN107578011A (en) | The decision method and device of key frame of video | |
| CN113191358A (en) | Metal part surface text detection method and system | |
| CN105184294A (en) | Inclination character judgment and identification method based on pixel tracking | |
| CN1161688C (en) | Character processing apparatus and method therefor | |
| CN109271882B (en) | Method for extracting color-distinguished handwritten Chinese characters | |
| CN114998217A (en) | Method for determining defect grade of glass substrate, computer device and storage medium | |
| CN1217292C (en) | Bill image face identification method | |
| CN107240185B (en) | A kind of crown word number identification method, device, equipment and storage medium | |
| CN1489745A (en) | Method for identifying image and device for realizing same | |
| CN114550173A (en) | Image preprocessing method and device, electronic equipment and readable storage medium | |
| CN110427909B (en) | Mobile terminal driving license detection method and system, electronic equipment and storage medium | |
| CN111611783A (en) | Positioning and dividing method and device for graphic table | |
| CN110807348A (en) | Method for removing interference lines in document image based on greedy algorithm | |
| CN110598575A (en) | Table layout analysis and extraction method and related device | |
| CN113537392B (en) | Method and device for identifying similar images, computing equipment and computer storage medium | |
| CN110674779B (en) | Feature fusion-based monocular far infrared pedestrian detection method | |
| CN114037700A (en) | Panel defect detection method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C19 | Lapse of patent right due to non-payment of the annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |