CN103729656A - Seal imprint identification method and identification system - Google Patents
Seal imprint identification method and identification system Download PDFInfo
- Publication number
- CN103729656A CN103729656A CN201410028632.9A CN201410028632A CN103729656A CN 103729656 A CN103729656 A CN 103729656A CN 201410028632 A CN201410028632 A CN 201410028632A CN 103729656 A CN103729656 A CN 103729656A
- Authority
- CN
- China
- Prior art keywords
- seal
- bill
- impression
- image
- word
- 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
Images
Landscapes
- Collating Specific Patterns (AREA)
Abstract
The invention discloses a seal imprint identification method and identification system. The seal imprint identification method comprises collecting a bill seal image and user information on a user bill; separating a bill seal imprint from the bill seal image, extracting a contour of the bill seal imprint and obtaining an image which only contains the bill seal imprint; computing moment invariant characteristic values and poor image characteristic values of the bill seal imprint, computing the relative Euclidean distance and image difference parameters of two imprints according to moment invariant characteristic values of a reserved seal imprint which is corresponding to the user information; computing image-text characteristic values of the bill seal imprint when the relative Euclidean distance and the image difference parameters are within a set threshold value range, comparing the image-text characteristic values of the bill seal imprint with image-text characteristic values of the reserved seal imprint separately and computing the similarity of the bill seal imprint and the reserved seal imprint; judging the bill seal imprint to be true when the similarity is within the set threshold value range.
Description
Technical field
The present invention relates to computer graphics disposal technology field, particularly about the comparison technology of computer picture, be specifically related to a kind of discrimination method and identification system of seal impression.
Background technology
According to the national relevant laws and regulations such as < < negotiable instrument law > >, < < payment method in settling account > > really booking according to the key element of validity, be the reserved specimen seal in bank to public client.Described reserved specimen seal is when account is opened an account, and is reserved in the payment mark of bank, checks mark on the respective pay voucher that enterprise submits to whether consistent with it in order to when payment.Therefore the application of image comparison technology is particularly important to banking industry.The traditional seal comparison method of bank is that the mode of identifying by artificial knuckle is compared the image of voucher stamped signature and the image of the reserved bank of client papery seal at present.Mainly there is following defect in this processing mode: traditional manual print test mode is completely by manually relying on naked eyes rule of thumb to judge, and the human factor such as experience, mood that is subject to appraiser affects greatlyr, therefore differentiate that precision is low, that seal is veritified speed is slow and have a very large operational risk.
Along with expanding economy, the business that seal is differentiated is more and more, and artificial discrimination method cannot meet the needs of real business.For realizing, improve discriminating speed, differentiate degree of accuracy and simple flow, occurred on the market at present that some adopt the mode of template matches, difference image or edge feature to realize the computer stamp checking system of seal authenticity, but above-mentioned computer stamp checking system is because the feature adopting is more single, and identification result is subject to the factor impacts such as noise and seal quality, therefore can not meet well the requirement of accuracy and reliability.
Summary of the invention
Goal of the invention of the present invention is discrimination method and the identification system of the seal impression that a kind of high accuracy and reliability are provided.
According to an aspect of the present invention, provide a kind of discrimination method of seal impression, having comprised:
Gather bill seal image and user profile on user's bill;
From the bill seal image gathering, isolate bill seal impression and carry out the profile extraction of bill seal impression, obtain the image that only comprises bill seal impression;
Calculate square invariant features value and the difference image eigenwert of bill seal impression, and according to the square invariant features value calculating bill seal impression of reserved specimen seal impression corresponding to this user profile and the relative Euclidean distance of reserved specimen seal impression, and according to the difference image eigenwert computed image difference parameter of bill seal impression and reserved specimen seal impression;
When Euclidean distance and image difference parameter are all within the scope of setting threshold relatively, calculate the picture and text eigenwert of bill seal impression, and the picture and text eigenwert of the picture and text eigenwert of bill seal impression and reserved specimen seal impression is compared one by one, according to comparison result, calculate the similarity of bill seal impression and reserved specimen seal impression;
Bill seal impression by similarity within the scope of setting threshold is defined as very.
Further, described discrimination method also comprises:
When Euclidean distance or difference image difference parameter be not within the scope of setting threshold relatively, adopt knuckle method of identification to differentiate bill seal impression.
Wherein, described knuckle method of identification comprises:
Described bill seal impression is divided into two parts by line of cut, and described line of cut cuts the important picture and text feature in profile and the profile of described bill seal impression;
By the Part I in described bill seal impression and identical with Part I/similar partly overlapping in described reserved specimen seal impression, and the profile outward flange of Part I is alignd with the profile outward flange of superimposed part in described reserved specimen seal impression;
According to the degree of registration of figure and word in the seal shape in figure and word and reserved specimen seal impression in the seal shape at cut place in described bill seal impression, seal hem width, seal, seal hem width, seal, draw differentiation result.
Preferably, described line of cut is straight line or folding line.
Further,, after the bill seal image gathering on user's bill, also comprise:
Described bill seal image is carried out to image pre-service.
Preferably, described image pre-service comprise to image carry out noise reduction, increase and decrease contrast and/or increase and decrease brightness.
Wherein, the impression of isolating bill seal from the bill seal image gathering comprises: by bill seal image by RGB color space conversion to HSV space and carry out image binaryzation, according to the binaryzation information separated of described bill seal image, go out the region identical/similar with bill seal impression colouring information, and this region is separation with other parts of bill seal image, to isolate the impression of bill seal.
The picture and text feature of described bill seal impression comprise following at least one:
Seal shape, seal radius, seal hem width, star angle is high, secondary star angle is high, star angle brachium, the physical size of seal impression central point and word bottom, the physical size at seal impression central point and word top, the word of seal word is high, the size of seal word bottom and seal scuncheon, the size in seal word top and seal frame outside, the font of seal word, the word space Angular Dimension of seal word, physical size between seal word bottom and seal frame peak, physical size between seal word bottom and seal frame minimum point, physical size between seal word top and seal frame peak, physical size between seal word top and seal frame minimum point, the physical size of Yu Xing angle, seal word top peak, the physical size of Yu Xing angle, seal word top brachium part, size between seal word bottom and star angle peak, size between seal word bottom and star angle brachium part, size between seal word peak and seal frame peak, size between seal word minimum point and seal frame peak, size between seal word peak and seal frame minimum point, size between seal word minimum point and seal frame minimum point, seal word word interval approaches the physical size at place and the physical size of seal word word interval farthest most.
According to another aspect of the present invention, also provide a kind of identification system of seal impression, having comprised:
Seal information collecting device, for gathering bill seal image and the user profile on user's bill;
Separation of images extraction element, for isolating bill seal impression from the bill seal image gathering and carrying out the profile extraction of bill seal impression, obtains the image that only comprises bill seal impression;
Image registration device, for calculating square invariant features value and the difference image eigenwert of bill seal impression, and according to the square invariant features value calculating bill seal impression of reserved specimen seal impression corresponding to this user profile and the relative Euclidean distance of reserved specimen seal impression, and according to the difference image eigenwert computed image difference parameter of bill seal impression and reserved specimen seal impression;
Image comparison device, for when Euclidean distance and image difference parameter are all within the scope of setting threshold relatively, calculate the picture and text eigenwert of bill seal impression, and the picture and text eigenwert of the picture and text eigenwert of bill seal impression and reserved specimen seal impression is compared one by one, according to comparison result, calculate the similarity of bill seal impression and reserved specimen seal impression; Bill seal impression by similarity within the scope of setting threshold is defined as very.
Further, the identification system of described seal impression also comprises:
Simulation knuckle recognition device, for when Euclidean distance or image difference parameter be not within the scope of setting threshold relatively, adopts knuckle method of identification to differentiate bill seal impression.
As shown from the above technical solution, seal impression discrimination method in the present invention, after the relative Euclidean distance and image difference parameter of being drawn a bill according to seal impression and reserved specimen seal impression by square invariant features, difference image feature calculation, also needs the picture and text features such as the figure of printing, word in seal further to compare.Hence one can see that, and the comparison process of the seal impression discrimination method in the present invention is more careful, can reach the object that improves seal discriminating accuracy and reliability.Prove by experiment, the percent of pass of the true seal of discrimination method of the present invention reaches 95%, and the discrimination of false seal approaches 100%.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, below will the accompanying drawing of required use in embodiment or description of the Prior Art be briefly described.Apparently, the accompanying drawing in below describing is only some embodiments of the present invention, for those of ordinary skills, can also obtain according to these accompanying drawing illustrated embodiments other embodiment and accompanying drawing thereof.
Fig. 1 shows the process flow diagram of the seal impression discrimination method that the embodiment of the present invention provides;
Fig. 2 shows the structural representation of seal impression identification system;
Fig. 3 shows the structural representation of image acquisition platform.
Embodiment
For making object of the present invention, technical scheme and advantage clearer, referring to accompanying drawing and enumerate preferred embodiment, the present invention is described in more detail.But, it should be noted that, many details of listing in instructions are only used to make reader to have a thorough understanding to one or more aspects of the present invention, even if do not have these specific details also can realize these aspects of the present invention.
The terms such as " module " used in this application, " system " are intended to comprise the entity relevant to computing machine, such as but not limited to hardware, firmware, combination thereof, software or executory software.For example, module can be, but be not limited in: thread, program and/or the computing machine of the process moved on processor, processor, object, executable program, execution.For instance, the application program of moving on computing equipment and this computing equipment can be modules.One or more modules can be positioned at an executory process and/or thread, and module also can and/or be distributed on a computing machine between two or more computing machines.
Fig. 1 shows the process flow diagram of the seal impression discrimination method that the embodiment of the present invention provides.As shown in Figure 1, specifically comprise following flow process.
S101: gather bill seal image and user profile on user's bill.
In concrete embodiment, by driving scanner, camera or the high image acquisition devices such as instrument of clapping to gather the bill seal image on user's bill.Bill seal image adopts unified specification to preserve.
Described user profile comprises the information such as seal card number, account, account name in an account book.
S102: the bill seal image collecting is carried out to image pre-service.
Preferably, the image pre-service of in this step, bill seal image being carried out comprises: adopt median filtering method to carry out noise reduction to bill seal image, and the bill seal image after noise reduction is increased to its contrast and brightness, so that bill seal image is more clear, strengthen the readability of bill seal image.
S103: isolate bill seal impression from the bill seal image gathering.
As the preferred embodiment in each embodiment, the concrete grammar of isolating the impression of bill seal in bill seal image is: by bill seal image by RGB color space conversion to HSV space and carry out image binaryzation, according to the binaryzation information separated of described bill seal image, go out the region identical/similar with bill seal impression colouring information, and this region is separation with other parts of bill seal image, to isolate the impression of bill seal.
S104: isolated bill seal impression is carried out to profile extraction, obtain the image that only comprises bill seal impression.
For being shaped as circular seal, by the circle detection algorithm based on Hough conversion, extract the profile of bill seal impression, to extract the image that only includes circular seal impression.
For being shaped as square seal, by the square detection algorithm based on Hough conversion, extract the profile of bill seal impression, to extract the image that only includes square seal impression.
Circle detection algorithm based on Hough conversion and the square detection algorithm based on Hough conversion are those skilled in the art of the present technique's known technology, are no longer described in detail herein.
S105: square invariant features value and the difference image eigenwert of calculating bill seal impression, and according to the relative Euclidean distance of two kinds of impressions of square invariant features value calculating of reserved specimen seal impression corresponding to this user profile, and according to the image difference parameter of two kinds of impressions of difference image eigenwert calculating of two kinds of impressions.
The square invariant features of figure refers to the square variate-value feature that image remains unchanged in the situations such as convergent-divergent, rotation, translation, comprises Hu square invariant, Zernike square invariant etc., and the present invention preferably adopts Hu square invariant to carry out image authentication.
Hu square invariant comprises seven square invariant features values.For the computing method of seven square invariant features values of Hu square invariant, thering is in the art its criterion calculation formula, repeat no more herein.
In the present invention, the extracting method of reserved specimen seal impression can be according to the method shown in S101～S105 step in the present invention, collecting after reserved specimen seal image, first reserved specimen seal image is carried out to image pre-service, then in reserved specimen seal image, isolate reserved specimen seal impression, again isolated bill seal impression is carried out to profile extraction, obtain the image that only comprises reserved specimen seal impression.According to square invariant features value and the difference image eigenwert of the image calculation reserved specimen seal impression of reserved specimen seal impression.
In addition, reserved specimen seal impression square invariant features value and difference image eigenwert also can adopt additive method draw and store, and when calculating the relative Euclidean distance of bill seal impression and reserved specimen seal impression and image difference parameter, call.
Calculating after seven square invariant features values of bill seal impression, and according to seven of reserved specimen seal impression square invariant features values, calculating the relative Euclidean distance of bill seal impression and reserved specimen seal impression.Bill seal impression is as follows with the computing formula of the relative Euclidean distance of reserved specimen seal impression:
In formula:
represent bill seal impression Hu square invariant features value,
represent reserved specimen seal impression Hu square invariant features value.
The computing method of the difference image of bill seal impression and reserved specimen seal impression are:
If MS is reserved specimen seal, SS is bill seal, and f (p) is the gray scale that on image, p is ordered, and the computing formula of the difference image eigenwert DIM of definition bill seal impression and reserved specimen seal impression is:
DIM=∑ [f
mS(p) ⊕ f
sS(p)]; (formula two)
Wherein, f
mS(p) be reserved specimen seal image, f
sS(p) be bill seal image, ⊕ is XOR operator.
Meanwhile, calculate and be present in reserved specimen seal and be not present in the set of the point in bill seal:
And: calculate and be present in bill seal and be not present in the set of the point in reserved specimen seal:
DIM, MDIM, tri-difference image eigenwerts of SDIM are carried out to medium filtering, obtain new difference image DIM ', MDIM ', SDIM ', add up three kinds of difference images area before and after medium filtering respectively, obtain area statistics amount ms, dim, mdim, sdim, dimf, mdimf, the sdimf corresponding with DIM, MDIM, SDIM, DIM ', MDIM ' and SDIM ' difference, according to these area statistics amounts, calculate the image difference parameter of bill seal impressions and reserved specimen seal impression.The computing formula of the image difference parameter of bill seal impression and reserved specimen seal impression is:
By setting the threshold value of parameter D and the threshold value of image difference parameter T, the true and false of bill seal is tentatively differentiated.When parameter D and image difference parameter T are all within the scope of setting threshold, enter step S106 bill seal is carried out to next step discriminating; Otherwise bill seal is considered as vacation, provides identification result.
S106: calculate the picture and text eigenwert of bill seal impression, and the picture and text eigenwert of the picture and text eigenwert of bill seal impression and reserved specimen seal impression is compared one by one, calculate the similarity of bill seal impression and reserved specimen seal impression according to comparison result.
Wherein, in the present invention the described picture and text feature of bill seal impression comprise following at least one: seal shape, seal radius, seal hem width, star angle is high, secondary star angle is high, star angle brachium, the physical size of seal impression central point and word bottom, the physical size at seal impression central point and word top, the word of seal word is high, the size of seal word bottom and seal scuncheon, the size in seal word top and seal frame outside, the font of seal word, the word space Angular Dimension of seal word, physical size between seal word bottom and seal frame peak, physical size between seal word bottom and seal frame minimum point, physical size between seal word top and seal frame peak, physical size between seal word top and seal frame minimum point, the physical size of Yu Xing angle, seal word top peak, the physical size of Yu Xing angle, seal word top brachium part, size between seal word bottom and star angle peak, size between seal word bottom and star angle brachium part, size between seal word peak and seal frame peak, size between seal word minimum point and seal frame peak, size between seal word peak and seal frame minimum point, size between seal word minimum point and seal frame minimum point, seal word interval approaches the physical size at place and the physical size of seal word word interval farthest most.
Preferably, in the present invention, the computing formula of the similarity s of bill seal impression and reserved specimen seal impression is:
S=(1-{ ∑ [p (i)-y (i)]/y (i) }/n) * 100% (formula six)
In formula: p (i) represents a certain picture and text eigenwert of bill seal impression, y (i) represents a certain picture and text eigenwert of reserved specimen seal impression, and n represents the number of the picture and text characteristic parameter that participates in similarity calculating.
Certainly, in formula six, about the computing formula of similarity, be exemplary, not calculate unique computing formula of similarity, those skilled in the art also can calculate similarity according to other computing formula, as long as can judge both degrees of approximation.
In the present embodiment, the threshold range of described similarity is different from the threshold range of parameter D and image difference parameter T, and its numerical range is 0～100%.
S107: judge that similarity is whether within the scope of setting threshold.
If similarity is within the scope of setting threshold, bill seal is true; Otherwise bill seal is false.
As the preferred embodiments of the invention, before computing parameter D and image difference parameter T, also seal image can be carried out to registration.Particularly, method for registering is: the centre of form of bill seal impression is overlapped the translation of bill seal impression with the centre of form of reserved specimen seal impression, then after rotating to an angle, realize registration with reserved specimen seal impression.Wherein, the anglec of rotation of bill seal impression is deflection angle poor of reserved specimen seal impression and bill seal impression.
Further, when the parameter D of bill seal impression and reserved specimen seal impression and image difference parameter T are not within the scope of setting threshold, adopt knuckle method of identification to differentiate bill seal impression.
Particularly, described knuckle method of identification comprises:
Described bill seal impression is divided into two parts by line of cut.By the Part I of described bill seal impression and in reserved specimen seal impression, identical to Part I/similar partly overlapping, and the profile outward flange of Part I is alignd with the profile outward flange of superimposed part in reserved specimen seal impression.
In the present embodiment, the Part I of bill seal impression is just used for the arbitrary part in two parts that the cut line of acute pyogenic infection of finger tip bill seal impression is divided into, is not intended to limit the specific part of bill seal impression after cut.
When the profile outward flange of the Part I of bill seal impression cannot align with the profile outward flange that carries out lap in reserved specimen seal impression, can preliminary judgement bill seal be false.
Further, when bill seal is tentatively defined as fictitious time, can be by selecting the line of cut different from line of cut before again to cut bill seal impression, bill seal impression is still divided into two parts, and by the Part I and in reserved specimen seal impression, identical to Part I/similar partly overlapping of bill seal impression after cutting, the outer peripheral degree of registration of profile of carrying out lap in the profile outward flange by Part I in bill seal impression and reserved specimen seal impression judges.
After the cutting times of setting, in bill seal impression, the profile outward flange of Part I still cannot align with the profile outward flange that carries out lap in reserved specimen seal impression, and bill seal is false.
After the profile outward flange of the Part I of bill seal impression can align with the profile outward flange that carries out lap in reserved specimen seal impression, according to the degree of registration of figure and word in the seal shape in figure and word and reserved specimen seal impression in the seal shape at cut place in described bill seal impression, seal hem width, seal, seal hem width, seal, draw differentiation result.
In the present embodiment, when differentiating the degree of registration of bill seal impression and reserved specimen seal impression, figure and word in seal shape, seal hem width, seal can be drawn to differentiation result after all comparing.
Or as another preferred version, when differentiating the degree of registration of bill seal impression and reserved specimen seal impression, can differentiate successively by the order of figure and word in seal shape, seal hem width, seal, cannot be on time in seal shape, in seal hem width and seal, figure and word are no longer differentiated.
In seal shape, seal hem width, seal, putting in order of figure and word is exemplary, and those skilled in the art can need to sort above-mentioned parameter according to it.
For differentiating accurately, during line of cut cutting bill seal impression, must cut apart the important picture and text feature in the profile of bill seal impression and profile.
In the present embodiment, the number of cut-off rule and shape are exemplary, and its number and shape can be set according to those skilled in the art's design requirement, and the present invention does not specifically limit its number and shape.
Preferably, the line of cut using in knuckle method of identification is preferably straight line or folding line.
As known from the above, seal impression discrimination method in the present invention, after the relative Euclidean distance and image difference parameter of being drawn a bill according to seal impression and reserved specimen seal impression by square invariant features, difference image feature calculation, is also further compared the picture and text features such as the figure of printing, word in seal.Prove by experiment, the percent of pass of the true seal of discrimination method of the present invention reaches 95%, and the discrimination of false seal approaches 100%.Hence one can see that, and the comparison process of the seal impression discrimination method in the present invention is more careful, can reach the object that improves seal discriminating accuracy and reliability.
In addition,, for the seal that adopts above-mentioned discrimination method to discern the false from the genuine, can adopt the method for simulation artificial cognition to differentiate, the drawback of avoiding Traditional Man knuckle recognition method to bring.
According to a further aspect in the invention, also provide a kind of seal impression identification system.Fig. 2 shows the structural representation of seal impression identification system.As shown in Figure 2, seal impression identification system comprises seal information collecting device 201, image segmenting device 202, image acquiring apparatus 203, image registration device 204 and image comparison device 205.
Seal information collecting device is for gathering bill seal image and the user profile on user's bill.In the present embodiment, seal information collecting device preferably adopts the image acquisition platform shown in Fig. 3, as shown in Figure 3, image acquisition platform comprises the flat base 31 for placing the bills such as check or file, the pillar 32 being connected perpendicular to flat base and with flat base, the top of pillar is connected with support 33, and image acquisition camera 34 is installed on support.
Preferably, pillar comprises upper supporting column and lower supporting rod two parts.Upper supporting column can move up and down and rotate with respect to lower supporting rod, so that image acquisition camera is adjusted.
Image segmenting device is for isolating bill seal impression at bill seal image.
Image acquiring apparatus, for isolated bill seal impression is carried out to profile extraction, obtains the image that only comprises bill seal impression.
Image registration device is for calculating square invariant features value and the difference image eigenwert of bill seal impression, and according to the relative Euclidean distance of two kinds of impressions of square invariant features value calculating of reserved specimen seal impression corresponding to this user profile, and according to the difference image eigenwert computed image difference parameter of two kinds of impressions.
Image comparison device is for when Euclidean distance and image difference parameter are all within the scope of setting threshold relatively, calculate the picture and text eigenwert of bill seal impression, and the picture and text eigenwert of the picture and text eigenwert of bill seal impression and reserved specimen seal impression is compared one by one, according to comparison result, calculate the similarity of bill seal impression and reserved specimen seal impression.Image comparison device provides differentiation result according to similarity, and by display device, such as display etc. shows differentiates result.。
Image segmenting device, image acquiring apparatus, image registration device and image comparison device can be integrated in be connected with image acquisition platform for carrying out the processor of image processing, also can be used as independent hardware unit and interconnect successively.
Further, the seal impression identification system in the present invention also comprises:
Simulation knuckle recognition device 206, for when Euclidean distance or image difference parameter be not within the scope of setting threshold relatively, adopts knuckle method of identification to differentiate bill seal impression.
Further, the seal impression identification system in the present invention also comprises: data storage device 207.
Data storage device is for storing view data, user profile data, the view data of various different bill seals and the system log message of storage seal impression identification system of reserved specimen seal.In embodiments of the present invention, the view data of reserved specimen seal impression also can be stored in remote server, calls in use.
Data storage device can be arranged in processor, also can be arranged at the remote server end communicating with processor.
The foregoing is only preferred embodiment of the present invention, not for limiting the scope of the invention.Within the spirit and principles in the present invention all, any modification of doing, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.
Claims (10)
1. a discrimination method for seal impression, is characterized in that, comprising:
Gather bill seal image and user profile on user's bill;
From the bill seal image gathering, isolate bill seal impression and carry out the profile extraction of bill seal impression, obtain the image that only comprises bill seal impression;
Calculate square invariant features value and the difference image eigenwert of bill seal impression, and according to the square invariant features value calculating bill seal impression of reserved specimen seal impression corresponding to this user profile and the relative Euclidean distance of reserved specimen seal impression, and according to the difference image eigenwert computed image difference parameter of bill seal impression and reserved specimen seal impression;
When Euclidean distance and image difference parameter are all within the scope of setting threshold relatively, calculate the picture and text eigenwert of bill seal impression, and the picture and text eigenwert of the picture and text eigenwert of bill seal impression and reserved specimen seal impression is compared one by one, according to comparison result, calculate the similarity of bill seal impression and reserved specimen seal impression;
Bill seal impression by similarity within the scope of setting threshold is defined as very.
2. discrimination method according to claim 1, is characterized in that, also comprises:
When Euclidean distance or difference image difference parameter be not within the scope of setting threshold relatively, adopt knuckle method of identification to differentiate bill seal impression.
3. discrimination method according to claim 2, is characterized in that, described knuckle method of identification comprises:
Described bill seal impression is divided into two parts by line of cut, and described line of cut cuts the important picture and text feature in profile and the profile of described bill seal impression;
By the Part I in described bill seal impression and identical with Part I/similar partly overlapping in described reserved specimen seal impression, and the profile outward flange of Part I is alignd with the profile outward flange of superimposed part in described reserved specimen seal impression;
According to the degree of registration of figure and word in the seal shape in figure and word and reserved specimen seal impression in the seal shape at cut place in described bill seal impression, seal hem width, seal, seal hem width, seal, draw differentiation result.
4. discrimination method according to claim 3, is characterized in that, described line of cut is straight line or folding line.
5. according to arbitrary described discrimination method in claim 1 to 4, it is characterized in that, after the bill seal image gathering on user's bill, also comprise:
Described bill seal image is carried out to image pre-service.
6. discrimination method according to claim 5, is characterized in that, described image pre-service comprises carries out noise reduction, increase and decrease contrast and/or increase and decrease brightness to image.
7. discrimination method according to claim 5, is characterized in that, the impression of isolating bill seal from the bill seal image gathering comprises:
By bill seal image by RGB color space conversion to HSV space and carry out image binaryzation, according to the binaryzation information separated of described bill seal image, go out the region identical/similar with bill seal impression colouring information, and this region is separation with other parts of bill seal image, to isolate the impression of bill seal.
8. discrimination method according to claim 1, is characterized in that, the picture and text feature of described bill seal impression comprise following at least one:
Seal shape, seal radius, seal hem width, star angle is high, secondary star angle is high, star angle brachium, the physical size of seal impression central point and word bottom, the physical size at seal impression central point and word top, the word of seal word is high, the size of seal word bottom and seal scuncheon, the size in seal word top and seal frame outside, the font of seal word, the word space Angular Dimension of seal word, physical size between seal word bottom and seal frame peak, physical size between seal word bottom and seal frame minimum point, physical size between seal word top and seal frame peak, physical size between seal word top and seal frame minimum point, the physical size of Yu Xing angle, seal word top peak, the physical size of Yu Xing angle, seal word top brachium part, size between seal word bottom and star angle peak, size between seal word bottom and star angle brachium part, size between seal word peak and seal frame peak, size between seal word minimum point and seal frame peak, size between seal word peak and seal frame minimum point, size between seal word minimum point and seal frame minimum point, seal word word interval approaches the physical size at place and the physical size of seal word word interval farthest most.
9. an identification system for seal impression, is characterized in that, comprising:
Seal information collecting device, for gathering bill seal image and the user profile on user's bill;
Separation of images extraction element, for isolating bill seal impression from the bill seal image gathering and carrying out the profile extraction of bill seal impression, obtains the image that only comprises bill seal impression;
Image registration device, for calculating square invariant features value and the difference image eigenwert of bill seal impression, and according to the square invariant features value calculating bill seal impression of reserved specimen seal impression corresponding to this user profile and the relative Euclidean distance of reserved specimen seal impression, and according to the difference image eigenwert computed image difference parameter of bill seal impression and reserved specimen seal impression;
Image comparison device, for when Euclidean distance and image difference parameter are all within the scope of setting threshold relatively, calculate the picture and text eigenwert of bill seal impression, and the picture and text eigenwert of the picture and text eigenwert of bill seal impression and reserved specimen seal impression is compared one by one, according to comparison result, calculate the similarity of bill seal impression and reserved specimen seal impression; Bill seal impression by similarity within the scope of setting threshold is defined as very.
10. the identification system of seal impression according to claim 9, is characterized in that, also comprises:
Simulation knuckle recognition device, for when Euclidean distance or image difference parameter be not within the scope of setting threshold relatively, adopts knuckle method of identification to differentiate bill seal impression.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410028632.9A CN103729656A (en) | 2014-01-22 | 2014-01-22 | Seal imprint identification method and identification system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410028632.9A CN103729656A (en) | 2014-01-22 | 2014-01-22 | Seal imprint identification method and identification system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103729656A true CN103729656A (en) | 2014-04-16 |
Family
ID=50453720
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410028632.9A Pending CN103729656A (en) | 2014-01-22 | 2014-01-22 | Seal imprint identification method and identification system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103729656A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107358184A (en) * | 2017-06-30 | 2017-11-17 | 中国科学院自动化研究所 | The extracting method and extraction element of document word |
| CN108010040A (en) * | 2017-12-20 | 2018-05-08 | 南通艾思达智能科技有限公司 | A kind of method for separating medical bill red seal |
| CN108197642A (en) * | 2017-12-25 | 2018-06-22 | 山东浪潮云服务信息科技有限公司 | A kind of seal discrimination method and device |
| WO2018166236A1 (en) * | 2017-03-13 | 2018-09-20 | 平安科技（深圳）有限公司 | Claim settlement bill recognition method, apparatus and device, and computer-readable storage medium |
| CN110378886A (en) * | 2019-07-22 | 2019-10-25 | 中国工商银行股份有限公司 | Image comparison method, image comparison device, electronic equipment and medium |
| CN110751139A (en) * | 2019-09-16 | 2020-02-04 | 深圳市国信合成科技有限公司 | Invoice seal identification method and device, computer equipment and storage medium |
| CN112466033A (en) * | 2020-12-15 | 2021-03-09 | 江苏国光信息产业股份有限公司 | Ultraviolet pattern based bill verification detection method and detection system thereof |
| CN114240780A (en) * | 2021-12-15 | 2022-03-25 | 中国建设银行股份有限公司 | Seal identification method, device, equipment, medium and product |
| CN114838670A (en) * | 2022-03-11 | 2022-08-02 | 南京北新智能科技有限公司 | Photovoltaic panel dust detection method based on color analysis |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1100827A (en) * | 1993-09-21 | 1995-03-29 | 北京大学 | Method and system for recognising seals |
| WO2001015382A1 (en) * | 1999-08-21 | 2001-03-01 | Kent Ridge Digital Labs | Legitimacy protection of electronic document and a printed copy thereof |
| CN101008985A (en) * | 2007-01-18 | 2007-08-01 | 章毅 | Seal identification system and controlling method thereof |
| CN101894268A (en) * | 2010-07-16 | 2010-11-24 | 西安理工大学 | Seal authenticity identification method capable of eliminating receipt interference |
-
2014
- 2014-01-22 CN CN201410028632.9A patent/CN103729656A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1100827A (en) * | 1993-09-21 | 1995-03-29 | 北京大学 | Method and system for recognising seals |
| WO2001015382A1 (en) * | 1999-08-21 | 2001-03-01 | Kent Ridge Digital Labs | Legitimacy protection of electronic document and a printed copy thereof |
| CN101008985A (en) * | 2007-01-18 | 2007-08-01 | 章毅 | Seal identification system and controlling method thereof |
| CN101894268A (en) * | 2010-07-16 | 2010-11-24 | 西安理工大学 | Seal authenticity identification method capable of eliminating receipt interference |
Non-Patent Citations (2)
| Title |
|---|
| 刘铁根 等: "基于嵌入式系统结构的印鉴鉴别系统", 《光学学报》 * |
| 郭磊民: "银行验印系统中的印鉴自动识别", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018166236A1 (en) * | 2017-03-13 | 2018-09-20 | 平安科技（深圳）有限公司 | Claim settlement bill recognition method, apparatus and device, and computer-readable storage medium |
| CN107358184A (en) * | 2017-06-30 | 2017-11-17 | 中国科学院自动化研究所 | The extracting method and extraction element of document word |
| CN108010040A (en) * | 2017-12-20 | 2018-05-08 | 南通艾思达智能科技有限公司 | A kind of method for separating medical bill red seal |
| CN108197642A (en) * | 2017-12-25 | 2018-06-22 | 山东浪潮云服务信息科技有限公司 | A kind of seal discrimination method and device |
| CN110378886A (en) * | 2019-07-22 | 2019-10-25 | 中国工商银行股份有限公司 | Image comparison method, image comparison device, electronic equipment and medium |
| CN110378886B (en) * | 2019-07-22 | 2021-09-10 | 中国工商银行股份有限公司 | Image comparison method, image comparison device, electronic device and medium |
| CN110751139A (en) * | 2019-09-16 | 2020-02-04 | 深圳市国信合成科技有限公司 | Invoice seal identification method and device, computer equipment and storage medium |
| CN112466033A (en) * | 2020-12-15 | 2021-03-09 | 江苏国光信息产业股份有限公司 | Ultraviolet pattern based bill verification detection method and detection system thereof |
| CN114240780A (en) * | 2021-12-15 | 2022-03-25 | 中国建设银行股份有限公司 | Seal identification method, device, equipment, medium and product |
| CN114838670A (en) * | 2022-03-11 | 2022-08-02 | 南京北新智能科技有限公司 | Photovoltaic panel dust detection method based on color analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103729656A (en) | Seal imprint identification method and identification system | |
| CN108717545B (en) | Bill identification method and system based on mobile phone photographing | |
| US7630520B2 (en) | Method and system for document comparison using cross plane comparison | |
| CN101059885B (en) | A ticket true/false verifying method | |
| CN105046807B (en) | A kind of counterfeit money recognition methods and system based on smart mobile phone | |
| JP2012008791A (en) | Form recognition device and form recognition method | |
| CN104766046A (en) | Detection and recognition algorithm conducted by means of traffic sign color and shape features | |
| Ahmed et al. | Forgery detection based on intrinsic document contents | |
| CN109492696B (en) | Seal authenticity identification method, device and system | |
| CN107180479A (en) | A kind of bill discrimination method, device, equipment and storage medium | |
| CN107067533A (en) | The method and device that a kind of bank note differentiates | |
| Abramova et al. | Detecting copy–move forgeries in scanned text documents | |
| CN111462388A (en) | Bill inspection method and device, terminal equipment and storage medium | |
| CN111368828A (en) | Multi-bill identification method and device | |
| CN112395995A (en) | Method and system for automatically filling and checking bill according to mobile financial bill | |
| Suresh et al. | Indian currency recognition and verification using image processing | |
| Zarin et al. | A hybrid fake banknote detection model using OCR, face recognition and hough features | |
| CN117058805A (en) | Banknote image processing method and system | |
| CN203882325U (en) | Seal impression discriminating system | |
| Ng et al. | An intelligent banknote recognition system by using machine learning with assistive technology for visually impaired people | |
| Sarkar et al. | Detecting counterfeit currency and identifying its source | |
| CN114926842A (en) | Dongba pictograph recognition method and device | |
| CN111291726B (en) | Medical bill sorting method, device, equipment and medium | |
| CN106846603A (en) | A kind of recognition methods of forge or true or paper money and its device | |
| CN107240184A (en) | A kind of plastics coin version knows method for distinguishing, device and equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20161214 Address after: 100070 Beijing, Fengtai District, South Fourth Ring Road West, No. two, building 188, floor 9 Applicant after: ChinaComm System Co., Ltd. Address before: 050081 Zhongshan West Road, Hebei, China, No. 589, No. Applicant before: Hebei FarEast Communication System Engineering Co., Ltd. |
|
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140416 |
|
| RJ01 | Rejection of invention patent application after publication |