CN104063859A - Method and system for detecting number of figures of account in figures in note image - Google Patents
Method and system for detecting number of figures of account in figures in note image Download PDFInfo
- Publication number
- CN104063859A CN104063859A CN201410040405.8A CN201410040405A CN104063859A CN 104063859 A CN104063859 A CN 104063859A CN 201410040405 A CN201410040405 A CN 201410040405A CN 104063859 A CN104063859 A CN 104063859A
- Authority
- CN
- China
- Prior art keywords
- bill
- figures
- bianry image
- amount
- right corner
- 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
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a method and a system for detecting the number of figures of an account in figures in a note image. The method comprises the steps of performing graying processing on an acquired note color image to obtain a note gray image, performing binarization processing on the note gray image, thereby obtaining a note binary image, performing convolution operation on the note binary image according to a preset top right color positioning template so as to position the region of the account in figures, and then extracting a binary image of the region of the note account in figures according to the positioning result, and processing the binary image of the region of the note account in figures by use of a statistical threshold method, thereby obtaining the number of figures of the note account in figures and a first character region. The method does not need manual operation so that the time of manually sorting notes and accidental errors caused by artificial operations are reduced; the method is short in time consumed and high in accuracy, and the efficiency of processing a batch of notes is improved. The method and the system can be widely applied to the image processing field.
Description
Technical field
The present invention relates to image processing field, especially a kind of detection method and system of bill images amount in figures figure place.
Background technology
For a long time, domestic bank to public service business, especially to accept be all that Receptionist by outlet completes to bill operation, need expend a large amount of human and material resources.Along with increasing of bank finance business, bill kind and quantity increase severely, and the batch robotization treatment technology of bill also arises at the historic moment.
According to the managerial demand of bank inside, self-service facility are not are not accepted more than 500,000 bills, and 50,000 yuan of following bills are directly uploaded, and 50,000 yuan of above bills need to carry out physical verification.Bank requires automatically to determine the mode of processing according to the division of the amount of the instrument: (1) national Check image exchange system, and requiring highest amount is 500,000; (2) People's Bank of China's wholesale system, requires amount to be greater than 50,000; (3) People's Bank of China's small amount system: require amount to be less than 50,000.
While carrying out amount of the instrument division, the definite of bill amount in figures figure place is the key process that the amount of the instrument is divided.Determining of bill amount in figures figure place still needs to adopt artificial mode at present, and this mode arranges the consuming time longer of bill, and the efficiency of batch processing is lower, and the accidental error that unavoidably manual operation brings, and degree of accuracy is not high.
Summary of the invention
In order to solve the problems of the technologies described above, the object of the invention is: the detection method that a kind of short, efficiency is high and degree of accuracy is high bill images amount in figures figure place consuming time is provided.
Another object of the present invention is: the detection system that a kind of short, efficiency is high and degree of accuracy is high bill images amount in figures figure place consuming time is provided.
The technical solution adopted for the present invention to solve the technical problems is: the detection method of bill images amount in figures figure place, comprising:
A, to gather bill coloured image carry out gray processing processing, thereby obtain bill gray level image;
B, bill gray level image is carried out to binary conversion treatment, thereby obtain bill bianry image;
C, according to default upper right corner locating template, bill bianry image is carried out to convolution algorithm, thereby the position in amount in figures region is positioned, then extract bill amount in figures region bianry image according to the result of location;
D, employing statistical threshold method are processed bill amount in figures region bianry image, thereby obtain figure place and the first character zone of bill amount in figures.
Further, described step C, it comprises:
C1, according to the wide high proportion of bill bianry image, determine the hunting zone in the bill bianry image rectangle frame upper right corner;
C2, the pixel in bill bianry image hunting zone and default upper right corner locating template are computed weighted, thereby the position in the bill bianry image rectangle frame upper right corner is accurately located;
C3, according to the position in the bill bianry image rectangle frame upper right corner, from bill bianry image, extract bill amount in figures region bianry image.
Further, described step C1, it comprises:
C11, the wide high proportion of bill bianry image is calculated, the computing formula of the wide high proportion r of described bill bianry image is: r=W/H, and in formula, W is the width of bill bianry image, H is the height of bill bianry image;
C12, according to the wide high proportion r of bill bianry image, whether decides bill has been torn counterfoil, if so, the Wei You upper left corner, the hunting zone [x in the bill bianry image rectangle frame upper right corner
0-a-c, y
0-b], the upper right corner [x
0+ a-c, y
0-b], the lower left corner [x
0-a-c, y
0+ b] and the lower right corner [x
0+ a-c, y
0+ b] this determined rectangular area, four rectangle frame summits; Otherwise, the Wei You upper left corner, the hunting zone [x in the bill bianry image rectangle frame upper right corner
0-a, y
0-b], the upper right corner [x
0+ a, y
0-b], the lower left corner [x
0-a, y
0+ b] and the lower right corner [x
0+ a, y
0+ b] this determined rectangular area, four rectangle frame summits; Wherein, x
0 Hey
0be respectively horizontal coordinate and the vertical coordinate of position, the complete bill rectangle frame upper right corner, a and b are respectively horizontal offset and the vertical offset of permission, and c is the counterfoil width tearing.
Further, described step C2, it is specially:
In hunting zone, from the upper left corner of bill bianry image, press from left to right, top-down order is translation upper right corner locating template successively, the center of upper right corner locating template is overlapped with the respective pixel point in hunting zone, then calculate the convolution value that upper right corner locating template and its cover bill bianry image pixel, the position in the described bill bianry image rectangle frame upper right corner is the corresponding position of convolution value maximal value calculating.
Further, described step C3, it is specially:
According to the position in the bill bianry image rectangle frame upper right corner, from bill bianry image, extract bill amount in figures region bianry image S[i, j], described bill amount in figures region bianry image S[i, j] computing formula be:
S[i,j]=D[x
0+i,y
0+j],
Wherein, i=1,2 ..., W
1, j=1,2 ..., H
1;w
1for the width of bill amount in figures region bianry image, H
1for the height of bill amount in figures region bianry image; S[i, j] represent that bill amount in figures region bianry image horizontal level is the value at the pixel place of i and upright position j, D[x
0+ i, y
0+ j] represent that bill bianry image horizontal level is x
0+ i and upright position are y
0the value at the pixel place of+j.
Further, described step D, it comprises:
D1, determine everybody position range of bill amount in figures according to the type of bill;
D2, from most significant digit, by the order from a high position to low level, in the position of everybody correspondence of bill amount in figures, to bill amount in figures, everybody carries out bill bianry image vertical direction foreground point and adds up, and the current statistical value of vertical direction foreground point and default statistical threshold are compared, if the current statistical value of vertical direction foreground point is less than default statistical threshold, forwards bill amount in figures next bit to and carry out bill bianry image vertical direction foreground point statistics; Otherwise present bit is the most significant digit that amount in figures is filled in, the most significant digit of then filling according to amount in figures is determined figure place and the first character zone of bill amount in figures.
Further, the type of described bill comprises check, electronics money order, pay-in slip, clearing application form, evidence for payment and credit voucher.
The present invention solves another technical scheme that its technical matters adopts: the detection system of bill images amount in figures figure place, comprising:
Gray space conversion module, for the bill coloured image gathering is carried out to gray processing processing, thereby obtains bill gray level image;
Binary conversion treatment module, for bill gray level image is carried out to binary conversion treatment, thereby obtains bill bianry image;
Bill amount in figures zone location and extraction module, for bill bianry image being carried out to convolution algorithm according to default upper right corner locating template, thereby the position to amount in figures region positions, then extract bill amount in figures region bianry image according to the result of location;
Bill amount in figures computing module, for adopting statistical threshold method to process bill amount in figures region bianry image, thereby obtains figure place and the first character zone of bill amount in figures;
The output terminal of described gray space conversion module is connected with extraction module and then with the input end of bill amount in figures computing module with bill amount in figures zone location by binary conversion treatment module successively.
Further, described bill amount in figures zone location and extraction module comprise:
Hunting zone determining unit, for according to the wide high proportion of bill bianry image, determines the hunting zone in the bill bianry image rectangle frame upper right corner;
Accurately positioning unit, for the pixel in bill bianry image hunting zone and default upper right corner locating template are computed weighted, thereby accurately locates the position in the bill bianry image rectangle frame upper right corner;
Extraction unit for according to the position in the bill bianry image rectangle frame upper right corner, extracts bill amount in figures region bianry image from bill bianry image;
The input end of described hunting zone determining unit is connected with the output terminal of binary conversion treatment module, and the output terminal of described hunting zone determining unit is connected with extraction unit and then with the input end of bill amount in figures computing module by accurate positioning unit successively.
The beneficial effect of method of the present invention is: by bill images being carried out to gray processing, binaryzation and convolution algorithm processing, amount in figures region in bill is extracted, and adopt statistical threshold method to determine the figure place of amount in figures, without manual operation, reduced the time of manual sorting bill and the accidental error that manual operation brings, consuming time short, degree of accuracy is high and promoted the efficiency of bill processing in batches.
The beneficial effect of system of the present invention is: by gray space conversion module, gray space conversion module and bill amount in figures zone location and extraction module, bill images is processed, extract the amount in figures region in bill, and adopt statistical threshold method to determine the figure place of amount in figures, without manual operation, reduced the time of manual sorting bill and the accidental error that manual operation brings, consuming time short, degree of accuracy is high and promoted the efficiency of bill processing in batches.
Brief description of the drawings
Below in conjunction with drawings and Examples, the invention will be further described.
Fig. 1 is the flow chart of steps of the detection method of bill images amount in figures figure place of the present invention;
Fig. 2 is the process flow diagram of step C of the present invention;
Fig. 3 is the process flow diagram of step C1 of the present invention;
Fig. 4 is the process flow diagram of step D of the present invention;
Fig. 5 is the functional schematic block diagram of the detection system of bill images amount in figures figure place of the present invention;
Fig. 6 is the structured flowchart of bill amount in figures zone location of the present invention and extraction module.
Embodiment
With reference to Fig. 1, the detection method of bill images amount in figures figure place, comprising:
A, to gather bill coloured image carry out gray processing processing, thereby obtain bill gray level image;
B, bill gray level image is carried out to binary conversion treatment, thereby obtain bill bianry image;
C, according to default upper right corner locating template, bill bianry image is carried out to convolution algorithm, thereby the position in amount in figures region is positioned, then extract bill amount in figures region bianry image according to the result of location;
D, employing statistical threshold method are processed bill amount in figures region bianry image, thereby obtain figure place and the first character zone of bill amount in figures.
Wherein, the bill coloured image gathering is carried out to gray processing processing, refer to the bill coloured image of collection is transformed to gray space by rgb space, the formula adopting when it converts is:
Y[i,j]=0.35R[i,j]+0.59G[i,j]+0.11B[i,j],
In formula, i=1,2 ..., W
0; J=1,2 ..., H
0;w
0for the width of bill coloured image, H
0for the height of bill coloured image, i represents the horizontal ordinate of bill coloured image pixel, and j represents the ordinate of bill coloured image pixel (i, j); R[i, j], G[i, j] and B[i, j] represent respectively red color component value, green component values and the blue component value of bill coloured image pixel (i, j), Y[i, j] represent the pixel value of bill gray level image pixel (i, j).
Bill gray level image is carried out to binary conversion treatment mainly to be referred to, bill gray level image is carried out to Threshold segmentation, make the pixel value of the pixel that is greater than segmentation threshold th equal 255(black), and make the pixel value of the pixel that is less than or equal to segmentation threshold th equal 0(white).
choosing of th will make the black font of bill gray level image and the interference that form data showed out and removed other information especially, when
the pixel value of pixel equals at 255 o'clock, and for extracting foreground target point, segmentation threshold th can get the value that 50(obtains according to priori), th also can obtain by adaptive threshold or OSTU threshold value calculation method.
Locating template refers to the each summit for locating bill bianry image rectangle frame, carries out the matrix of convolution algorithm (ranking operation) with each summit of bill bianry image rectangle frame.Upper right corner locating template is for locating the summit, the upper right corner of bill bianry image rectangle frame.
With reference to Fig. 2, be further used as preferred embodiment, described step C, it comprises:
C1, according to the wide high proportion of bill bianry image, determine the hunting zone in the bill bianry image rectangle frame upper right corner;
C2, the pixel in bill bianry image hunting zone and default upper right corner locating template are computed weighted, thereby the position in the bill bianry image rectangle frame upper right corner is accurately located;
C3, according to the position in the bill bianry image rectangle frame upper right corner, from bill bianry image, extract bill amount in figures region bianry image.
With reference to Fig. 3, be further used as preferred embodiment, described step C1, it comprises:
C11, the wide high proportion of bill bianry image is calculated, the computing formula of the wide high proportion r of described bill bianry image is: r=W/H, and in formula, W is the width of bill bianry image, H is the height of bill bianry image;
C12, according to the wide high proportion r of bill bianry image, whether decides bill has been torn counterfoil, if so, the Wei You upper left corner, the hunting zone [x in the bill bianry image rectangle frame upper right corner
0-a-c, y
0-b], the upper right corner [x
0+ a-c, y
0-b], the lower left corner [x
0-a-c, y
0+ b] and the lower right corner [x
0+ a-c, y
0+ b] this determined rectangular area, four rectangle frame summits; Otherwise, the Wei You upper left corner, the hunting zone [x in the bill bianry image rectangle frame upper right corner
0-a, y
0-b], the upper right corner [x
0+ a, y
0-b], the lower left corner [x
0-a, y
0+ b] and the lower right corner [x
0+ a, y
0+ b] this determined rectangular area, four rectangle frame summits; Wherein, x
0 Hey
0be respectively horizontal coordinate and the vertical coordinate of position, the complete bill rectangle frame upper right corner, a and b are respectively horizontal offset and the vertical offset of permission, and c is the counterfoil width tearing.
Taking check as example, if when the wide high proportion r > 2.5 calculating, show that check is complete check; Otherwise, show that check is the check that tears counterfoil.
Be further used as preferred embodiment, described step C2, it is specially:
In hunting zone, from the upper left corner of bill bianry image, press from left to right, top-down order is translation upper right corner locating template successively, the center of upper right corner locating template is overlapped with the respective pixel point in hunting zone, then calculate the convolution value that upper right corner locating template and its cover bill bianry image pixel, the position in the described bill bianry image rectangle frame upper right corner is the corresponding position of convolution value maximal value calculating.
With T[m, n] represent the pixel value of the capable m row of upper right corner locating template n, T[m, n] meet:
Taking check as example:
In the time of r > 2.5, each pixel and upper right corner locating template T[m in hunting zone, the bill bianry image rectangle frame upper right corner, n] convolution value t[i, j] computing method as follows:
,
Wherein, i=x
0-a, x
0-a+1 ..., x
0+ a; j=y
0-b, y
0-b+1 ..., y
0+ b; D[i, j] represent that bill bianry image horizontal level is the value at j and the upright position pixel place (i, j) that is j.
In the time of r≤2.5, each pixel and upper right corner locating template T[m in hunting zone, the bill bianry image rectangle frame upper right corner, n] convolution value t[i, j] computing method as follows:
Wherein, i=x
0-a-c, x
0-a-c+1 ..., x
0+ a-c; j=y
0-b, y
0-b+1 ..., y
0+ b; D[i, j] represent that bill bianry image horizontal level is the value at j and the upright position pixel place (i, j) that is j.
The convolution value t[i of each pixel in the hunting zone, the bill bianry image rectangle frame upper right corner calculating, j] in, t[i, j] the corresponding position of maximal value is just the position in the bill bianry image rectangle frame upper right corner.
Be further used as preferred embodiment, described step C3, it is specially:
According to the position in the bill bianry image rectangle frame upper right corner, from bill bianry image, extract bill amount in figures region bianry image S[i, j], described bill amount in figures region bianry image S[i, j] computing formula be:
S[i,j]=D[x
0+i,y
0+j],
Wherein, i=1,2 ..., W
1, j=1,2 ..., H
1;w
1for the width of bill amount in figures region bianry image, H
1for the height of bill amount in figures region bianry image; S[i, j] represent that bill amount in figures region bianry image horizontal level is the value at the pixel place of i and upright position j, D[x
0+ i, y
0+ j] represent that bill bianry image horizontal level is x
0+ i and upright position are y
0the value at the pixel place of+j.
The position in the bill bianry image rectangle frame upper right corner is just the position in the amount in figures upper right corner, because the relative position in the position in the upper left corner, amount in figures region and the bill bianry image rectangle frame upper right corner is fixed, so obtain after the relative position in the bill bianry image rectangle frame upper right corner, amount in figures region being extracted from whole bill bianry image, thereby obtain bill amount in figures region bianry image.
Taking check as example, 200dpi(dpi is the abbreviation of " Dot Per Inch ", be counting or line number of printing of per inch, be used for representing printer resolution, being usually said hardware resolution, is a large standard of weighing printer precision, in general, this value is larger, shows that the precision of printer is higher) width W of check amount in figures region bianry image
1the height H of=341,200dpi check amount in figures region bianry image
1=70.
With reference to Fig. 4, be further used as preferred embodiment, described step D, it comprises:
D1, determine everybody position range of bill amount in figures according to the type of bill;
D2, from most significant digit, by the order from a high position to low level, in the position of everybody correspondence of bill amount in figures, to bill amount in figures, everybody carries out bill bianry image vertical direction foreground point and adds up, and the current statistical value of vertical direction foreground point and default statistical threshold are compared, if the current statistical value of vertical direction foreground point is less than default statistical threshold, forwards bill amount in figures next bit to and carry out bill bianry image vertical direction foreground point statistics; Otherwise present bit is the most significant digit that amount in figures is filled in, the most significant digit of then filling according to amount in figures is determined figure place and the first character zone of bill amount in figures.
In bill, totally 11, the bill amount in figures region of check, remittance voucher, pay-in slip, clearing application form, totally 10 of evidence for payment and credit vouchers, therefore, first want the type of decides bill.
The present embodiment describes as an example of definite method of 200dpi check amount in figures figure place example.
200dpi check is turned left from the right side and is followed successively by the 1st, and the 2nd, the 3rd, the 4th, the 5th, the 6th, the 7th, the 8th, the 9th, the 10th, the 11st.If its everybody band of position horizontal extent is followed successively by the 1st (310~340), the 2nd (279~309), the 3rd (248~278), the 4th (217~247), the 5th (186~216), the 6th (155~185), the 7th (124~154), the 8th (93~123), the 9th (62~92), the 10th (31~61), the 11st (30~0).Vertical range is (30~70).
Obtaining after everybody position range, by order from left to right, from most significant digit, in the position of everybody correspondence of amount in figures, carry out bill amount in figures bianry image vertical direction foreground point statistics, obtain the statistical value of the vertical direction foreground point of every corresponding position:
As k(k=1,2,3,, 10,11) and the vertical direction foreground point system of position is less than statistical threshold TH, represent that this is without fill substance, proceed scanning from this beginning toward the right, until p(p=1,2,3 ..., 10,11) the vertical direction foreground point statistical value in corresponding region, position is greater than TH, the figure place that now figure place P just fills in for amount in figures.The figure place of the present embodiment indication refers to the figure place that has comprised " $ ", for example scanning from left to right, the vertical direction foreground point statistical value of 5 is all less than TH above, in kilobit, the vertical direction foreground point statistical value of the 6th is greater than TH, amount in figures is the order of magnitude hundred (the 1st correspond to point), and amount in figures adds that " $ " figure place is altogether 6.Meanwhile, in the 5th region, corresponding foreground area extracts as the first character zone.
The TH of the present embodiment is statistical threshold, and the value of TH is less than everybody pixel number of the amount of money of filling in bill amount in figures bianry image.The bill of the present embodiment is 200dpi bill, so desirable TH=93.
Be further used as preferred embodiment, the type of described bill comprises check, electronics money order, pay-in slip, clearing application form, evidence for payment and credit voucher.
With reference to Fig. 5, the detection system of bill images amount in figures figure place, comprising:
Gray space conversion module, for the bill coloured image gathering is carried out to gray processing processing, thereby obtains bill gray level image;
Binary conversion treatment module, for bill gray level image is carried out to binary conversion treatment, thereby obtains bill bianry image;
Bill amount in figures zone location and extraction module, for bill bianry image being carried out to convolution algorithm according to default upper right corner locating template, thereby the position to amount in figures region positions, then extract bill amount in figures region bianry image according to the result of location;
Bill amount in figures computing module, for adopting statistical threshold method to process bill amount in figures region bianry image, thereby obtains figure place and the first character zone of bill amount in figures;
The output terminal of described gray space conversion module is connected with extraction module and then with the input end of bill amount in figures computing module with bill amount in figures zone location by binary conversion treatment module successively.
With reference to Fig. 6, be further used as preferred embodiment, described bill amount in figures zone location and extraction module comprise:
Hunting zone determining unit, for according to the wide high proportion of bill bianry image, determines the hunting zone in the bill bianry image rectangle frame upper right corner;
Accurately positioning unit, for the pixel in bill bianry image hunting zone and default upper right corner locating template are computed weighted, thereby accurately locates the position in the bill bianry image rectangle frame upper right corner;
Extraction unit for according to the position in the bill bianry image rectangle frame upper right corner, extracts bill amount in figures region bianry image from bill bianry image;
The input end of described hunting zone determining unit is connected with the output terminal of binary conversion treatment module, and the output terminal of described hunting zone determining unit is connected with extraction unit and then with the input end of bill amount in figures computing module by accurate positioning unit successively.
Compared with needs technology, the present invention is by carrying out gray processing, binaryzation and convolution algorithm processing to bill images, amount in figures region in bill is extracted, and adopt statistical threshold method to determine the figure place of amount in figures, without manual operation, reduced the time of manual sorting bill and the accidental error that manual operation brings, consuming time short, degree of accuracy is high and promoted the efficiency of bill processing in batches.
More than that better enforcement of the present invention is illustrated, but the invention is not limited to described embodiment, those of ordinary skill in the art also can make all equivalent variations or replacement under the prerequisite without prejudice to spirit of the present invention, and the distortion that these are equal to or replacement are all included in the application's claim limited range.
Claims (9)
1. the detection method of bill images amount in figures figure place, is characterized in that: comprising:
A, to gather bill coloured image carry out gray processing processing, thereby obtain bill gray level image;
B, bill gray level image is carried out to binary conversion treatment, thereby obtain bill bianry image;
C, according to default upper right corner locating template, bill bianry image is carried out to convolution algorithm, thereby the position in amount in figures region is positioned, then extract bill amount in figures region bianry image according to the result of location;
D, employing statistical threshold method are processed bill amount in figures region bianry image, thereby obtain figure place and the first character zone of bill amount in figures.
2. the detection method of bill images amount in figures figure place according to claim 1, is characterized in that: described step C, and it comprises:
C1, according to the wide high proportion of bill bianry image, determine the hunting zone in the bill bianry image rectangle frame upper right corner;
C2, the pixel in bill bianry image hunting zone and default upper right corner locating template are computed weighted, thereby the position in the bill bianry image rectangle frame upper right corner is accurately located;
C3, according to the position in the bill bianry image rectangle frame upper right corner, from bill bianry image, extract bill amount in figures region bianry image.
3. the detection method of bill images amount in figures figure place according to claim 2, is characterized in that: described step C1, and it comprises:
C11, the wide high proportion of bill bianry image is calculated, the computing formula of the wide high proportion r of described bill bianry image is: r=W/H, and in formula, W is the width of bill bianry image, H is the height of bill bianry image;
C12, according to the wide high proportion r of bill bianry image, whether decides bill has been torn counterfoil, if so, the Wei You upper left corner, the hunting zone [x in the bill bianry image rectangle frame upper right corner
0-a-c, y
0-b], the upper right corner [x
0+ a-c, y
0-b], the lower left corner [x
0-a-c, y
0+ b] and the lower right corner [x
0+ a-c, y
0+ b] this determined rectangular area, four rectangle frame summits; Otherwise, the Wei You upper left corner, the hunting zone [x in the bill bianry image rectangle frame upper right corner
0-a, y
0-b], the upper right corner [x
0+ a, y
0-b], the lower left corner [x
0-a, y
0+ b] and the lower right corner [x
0+ a, y
0+ b] this determined rectangular area, four rectangle frame summits; Wherein, x
0 Hey
0be respectively horizontal coordinate and the vertical coordinate of position, the complete bill rectangle frame upper right corner, a and b are respectively horizontal offset and the vertical offset of permission, and c is the counterfoil width tearing.
4. the detection method of bill images amount in figures figure place according to claim 3, is characterized in that: described step C2, and it is specially:
In hunting zone, from the upper left corner of bill bianry image, press from left to right, top-down order is translation upper right corner locating template successively, the center of upper right corner locating template is overlapped with the respective pixel point in hunting zone, then calculate the convolution value that upper right corner locating template and its cover bill bianry image pixel, the position in the described bill bianry image rectangle frame upper right corner is the corresponding position of convolution value maximal value calculating.
5. the detection method of bill images amount in figures figure place according to claim 4, is characterized in that: described step C3, and it is specially:
According to the position in the bill bianry image rectangle frame upper right corner, from bill bianry image, extract bill amount in figures region bianry image S[i, j], described bill amount in figures region bianry image S[i, j] computing formula be:
S[i,j]=D[x
0+i,y
0+j],
Wherein, i=1,2 ..., W
1, j=1,2 ..., H
1;w
1for the width of bill amount in figures region bianry image, H
1for the height of bill amount in figures region bianry image; S[i, j] represent that bill amount in figures region bianry image horizontal level is the value at the pixel place of i and upright position j, D[x
0+ i, y
0+ j] represent that bill bianry image horizontal level is x
0+ i and upright position are y
0the value at the pixel place of+j.
6. the detection method of bill images amount in figures figure place according to claim 5, is characterized in that: described step D, and it comprises:
D1, determine everybody position range of bill amount in figures according to the type of bill;
D2, from most significant digit, by the order from a high position to low level, in the position of everybody correspondence of bill amount in figures, to bill amount in figures, everybody carries out bill bianry image vertical direction foreground point and adds up, and the current statistical value of vertical direction foreground point and default statistical threshold are compared, if the current statistical value of vertical direction foreground point is less than default statistical threshold, forwards bill amount in figures next bit to and carry out bill bianry image vertical direction foreground point statistics; Otherwise present bit is the most significant digit that amount in figures is filled in, the most significant digit of then filling according to amount in figures is determined figure place and the first character zone of bill amount in figures.
7. the detection method of bill images amount in figures figure place according to claim 6, is characterized in that: the type of described bill comprises check, electronics money order, pay-in slip, clearing application form, evidence for payment and credit voucher.
8. the detection system of bill images amount in figures figure place, is characterized in that: comprising:
Gray space conversion module, for the bill coloured image gathering is carried out to gray processing processing, thereby obtains bill gray level image;
Binary conversion treatment module, for bill gray level image is carried out to binary conversion treatment, thereby obtains bill bianry image;
Bill amount in figures zone location and extraction module, for bill bianry image being carried out to convolution algorithm according to default upper right corner locating template, thereby the position to amount in figures region positions, then extract bill amount in figures region bianry image according to the result of location;
Bill amount in figures computing module, for adopting statistical threshold method to process bill amount in figures region bianry image, thereby obtains figure place and the first character zone of bill amount in figures;
The output terminal of described gray space conversion module is connected with extraction module and then with the input end of bill amount in figures computing module with bill amount in figures zone location by binary conversion treatment module successively.
9. the detection system of bill images amount in figures figure place according to claim 8, is characterized in that: described bill amount in figures zone location and extraction module comprise:
Hunting zone determining unit, for according to the wide high proportion of bill bianry image, determines the hunting zone in the bill bianry image rectangle frame upper right corner;
Accurately positioning unit, for the pixel in bill bianry image hunting zone and default upper right corner locating template are computed weighted, thereby accurately locates the position in the bill bianry image rectangle frame upper right corner;
Extraction unit for according to the position in the bill bianry image rectangle frame upper right corner, extracts bill amount in figures region bianry image from bill bianry image;
The input end of described hunting zone determining unit is connected with the output terminal of binary conversion treatment module, and the output terminal of described hunting zone determining unit is connected with extraction unit and then with the input end of bill amount in figures computing module by accurate positioning unit successively.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410040405.8A CN104063859A (en) | 2014-01-27 | 2014-01-27 | Method and system for detecting number of figures of account in figures in note image |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410040405.8A CN104063859A (en) | 2014-01-27 | 2014-01-27 | Method and system for detecting number of figures of account in figures in note image |
Publications (1)
Publication Number | Publication Date |
---|---|
CN104063859A true CN104063859A (en) | 2014-09-24 |
Family
ID=51551551
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410040405.8A Pending CN104063859A (en) | 2014-01-27 | 2014-01-27 | Method and system for detecting number of figures of account in figures in note image |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104063859A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104952143A (en) * | 2015-06-18 | 2015-09-30 | 辰通智能设备(深圳)有限公司 | Color change detection method and color change detection system for red waterlines of bill images |
CN105005969A (en) * | 2015-06-18 | 2015-10-28 | 辰通智能设备(深圳)有限公司 | Note image correction detecting method and system |
CN110969408A (en) * | 2019-11-08 | 2020-04-07 | 国网辽宁省电力有限公司 | Material settlement full-process integrated management platform generation system and method |
CN111027532A (en) * | 2019-12-11 | 2020-04-17 | 上海眼控科技股份有限公司 | System and method for identifying tax amount of insurance policy vehicle and ship for forced insurance |
-
2014
- 2014-01-27 CN CN201410040405.8A patent/CN104063859A/en active Pending
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104952143A (en) * | 2015-06-18 | 2015-09-30 | 辰通智能设备(深圳)有限公司 | Color change detection method and color change detection system for red waterlines of bill images |
CN105005969A (en) * | 2015-06-18 | 2015-10-28 | 辰通智能设备(深圳)有限公司 | Note image correction detecting method and system |
CN105005969B (en) * | 2015-06-18 | 2018-04-17 | 深圳辰通智能股份有限公司 | A kind of bill images alter detection method and system |
CN110969408A (en) * | 2019-11-08 | 2020-04-07 | 国网辽宁省电力有限公司 | Material settlement full-process integrated management platform generation system and method |
CN110969408B (en) * | 2019-11-08 | 2024-04-05 | 国网辽宁省电力有限公司 | Material settlement whole-flow integrated management platform generation system and method |
CN111027532A (en) * | 2019-12-11 | 2020-04-17 | 上海眼控科技股份有限公司 | System and method for identifying tax amount of insurance policy vehicle and ship for forced insurance |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103824373B (en) | A kind of bill images amount of money sorting technique and system | |
CN107633239B (en) | Bill classification and bill field extraction method based on deep learning and OCR | |
CN109657665B (en) | Invoice batch automatic identification system based on deep learning | |
US10430681B2 (en) | Character segmentation and recognition method | |
US20070253040A1 (en) | Color scanning to enhance bitonal image | |
CN102567764A (en) | Bill certificate and system for improving electronic image recognition efficiency | |
CN104966348B (en) | A kind of bill images key element integrality detection method and system | |
US20170308768A1 (en) | Character information recognition method based on image processing | |
CN104134072A (en) | Answer sheet identification method | |
CN111178355B (en) | Seal identification method, device and storage medium | |
CN104063859A (en) | Method and system for detecting number of figures of account in figures in note image | |
CN104952077B (en) | A kind of bill images greasy dirt detection method and system | |
CN105005969B (en) | A kind of bill images alter detection method and system | |
CN102136058A (en) | Bar code image identification method | |
CN103886612A (en) | Automatic water level extraction method and system based on reservoir monitoring camera | |
CN101887521A (en) | Method and terminal for rectifying deviation of file | |
CN102855232A (en) | Table analysis and edit processing method | |
CN108108753A (en) | A kind of recognition methods of check box selection state based on support vector machines and device | |
CN111680690A (en) | Character recognition method and device | |
CN105160682A (en) | Image edge detection method and apparatus | |
CN109598271A (en) | A kind of character segmentation method and device | |
CN113591866A (en) | Special job certificate detection method and system based on DB and CRNN | |
CN107122775A (en) | A kind of Android mobile phone identity card character identifying method of feature based matching | |
CN107133647A (en) | A kind of quick Manuscripted Characters Identification Method | |
CN106709474A (en) | Handwritten telephone number identification, verification and information sending system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
CB02 | Change of applicant information |
Address after: Shenzhen Nanshan District City, Guangdong province 518000 white pine Luxili Nangang first industrial park, building seven floor 1-3 Applicant after: TOUCHSTONE INTELLIGENCE SYSTEM (SHENZHEN) CO., LTD. Address before: Shenzhen Nanshan District City, Guangdong province 518000 white pine Luxili Nangang first Industrial Park 7 Applicant before: Automated Touchstone Machines Limited |
|
COR | Change of bibliographic data | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140924 |
|
RJ01 | Rejection of invention patent application after publication |