CN108961200A - A kind of dust detection method and device - Google Patents
A kind of dust detection method and device Download PDFInfo
- Publication number
- CN108961200A CN108961200A CN201710347931.2A CN201710347931A CN108961200A CN 108961200 A CN108961200 A CN 108961200A CN 201710347931 A CN201710347931 A CN 201710347931A CN 108961200 A CN108961200 A CN 108961200A
- Authority
- CN
- China
- Prior art keywords
- image
- dust
- pixel
- width
- area
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/94—Investigating contamination, e.g. dust
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
- G07D7/00—Testing specially adapted to determine the identity or genuineness of valuable papers or for segregating those which are unacceptable, e.g. banknotes that are alien to a currency
- G07D7/181—Testing mechanical properties or condition, e.g. wear or tear
- G07D7/187—Detecting defacement or contamination, e.g. dirt
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
Abstract
The embodiment of the invention discloses a kind of dust detection method and device, and wherein method includes: to obtain area to be tested image；Determine in the area to be tested image whether there is dust image；If so, determining the track width of the dust image；If the track width of the dust image is greater than predetermined width threshold value, generates dedusting and remind instruction and show.The embodiment of the present invention solves finance device dedusting caused by artificial determining dedusting time not in time, the problem of reducing paper money recognition accuracy or staff's progress idle work, realize the dust deposit state inside automatic checkout equipment, the paper money recognition accuracy for improving finance device, improves the working efficiency of staff.
Description
Technical field
The present embodiments relate to image processing techniques more particularly to a kind of dust detection method and device.
Background technique
With the continuous development of city intelligent, more and more financial self-service equipments such as ATM (Automatic Teller
Machine, ATM) or CRS (Cash Recycling System, self-service automatic teller machine) etc. come into operation.
After above-mentioned finance device operation a period of time, the meeting inside equipment generates dust deposit, and dust will affect equipment
It operates normally, especially the dust inside finance device currency checking module can introduce loud noise during detecting forge or true or paper money,
Cause paper money recognition discrimination unstable.
Currently, be usually that staff periodically carries out according to working experience to the dedusting of finance device, regular dedusting time
It is equally empirically determined by staff, it easily leads to staff's dedusting not in time, reduces paper money recognition correctness, Huo Zhe
The problem of increasing the workload of staff in the case where dedusting is not needed.
Summary of the invention
The present invention provides a kind of dust detection method and device, to realize automatic realization equipment internal dust detection.
In a first aspect, the embodiment of the invention provides a kind of dust detection methods, this method comprises:
Obtain area to be tested image；
Determine in the area to be tested image whether there is dust image；
If so, determining the track width of the dust image；
If the track width of the dust image is greater than predetermined width threshold value, generates dedusting and remind instruction and show.
Further, area to be tested image is obtained, comprising:
Obtain the gray level image of transmission channel；
According to the relative position information of idler wheel and the transmission channel, the idler wheel picture position of the gray level image is determined,
The idler wheel is located in the transmission channel；
The area to be tested image is intercepted in the gray level image of the transmission channel according to the idler wheel picture position,
The area to be tested image includes background image and banknote image, and the banknote image is located in the background image region
Portion.
Further, determine in the area to be tested image whether there is dust image, comprising:
Determine in the background image the accumulative difference of the pixel value of the first default row pixel and；
If the accumulative difference and being greater than default difference and threshold value, it is determined that there is dust image in the background image.
Further, the track width of the dust image is determined, comprising:
Obtain the first pixel value of each neighbor pixel in the second default row pixel of described image to be detected；
The first maximum value and first time maximum difference in first pixel value are screened, and determination described first is most
It is big to be worth pixel corresponding with the first time maximum value；
The first distance of first maximum value described in each row pixel and the corresponding pixel of the first time maximum value is true
It is set to the dust width of each row pixel；
The mean value of the dust width is determined as to the track width of the dust image.
Further, determine in the area to be tested image whether there is dust image, comprising:
Silhouette markup is carried out to the banknote image, and determines the connected domain quantity of the banknote image；
If the connected domain quantity is greater than 1, it is determined that there is dust image in the banknote image.
Further, the track width of the dust image is determined, comprising:
Determine the second pixel value of neighbor pixel in the default row pixel of third in adjacent connected area segmentation region；
The second maximum value and second of maximum value in second pixel value are screened, and determines that described second is maximum
It is worth pixel corresponding with second of maximum value；
Determine the second distance of second maximum value and the corresponding pixel of second of maximum value；
The sum of the second distance that row pixel is corresponded in each adjacent connected area segmentation region is determined as the row
The dust width of pixel；
The mean value of the dust width is determined as to the track width of the dust image.
Further, before the mean value of the dust width is determined as the track width of the dust image, comprising:
The dust width is subjected to numerical ordering；
Correspondingly, the mean value of the dust width to be determined as to the track width of the dust image, comprising:
The dust width that predetermined order range is screened in the numerical ordering, the mean value of the dust width after screening is made
For the track width of the dust image.
Second aspect, the embodiment of the invention also provides a kind of dust detection, which includes:
Image collection module, for obtaining area to be tested image；
Dust image determining module, for determining in the area to be tested image whether there is dust image；
Track width determining module, if for having dust image in the area to be tested image, it is determined that the dust
The track width of image；
Directive generation module is reminded in dedusting, if the track width for the dust image is greater than predetermined width threshold value,
Dedusting is generated to remind instruction and show.
Further, described image acquisition module includes:
Transmission channel image acquisition unit, for obtaining the gray level image of transmission channel；
Idler wheel picture position determination unit determines institute for the relative position information according to idler wheel and the transmission channel
The idler wheel picture position of gray level image is stated, the idler wheel is located in the transmission channel；
Image interception unit, for intercepting institute in the gray level image of the transmission channel according to the idler wheel picture position
Area to be tested image is stated, the area to be tested image includes background image and banknote image, and the banknote image is located at institute
It states inside background image region.
Further, the dust image determining module includes:
Accumulative difference and determination unit, for determining the accumulative of the pixel value of the first default row pixel in the background image
Difference and；
First dust image determination unit, if for the accumulative difference and being greater than default difference and threshold value, it is determined that institute
Stating in background image has dust image.
Further, the track width determining module includes:
First difference true ground unit, each neighbor pixel in the second default row pixel for obtaining described image to be detected
The first pixel value；
First pixel determination unit, for screening the first maximum value in first pixel value and first time most
Big difference, and determine first maximum value and the corresponding pixel of the first time maximum value；
First dust width determination unit is used for the first maximum value described in each row pixel and the first time maximum value
The first distance of corresponding pixel is determined as the dust width of each row pixel；
First track width determination unit, for the mean value of the dust width to be determined as to the track of the dust image
Width.
Further, the dust image determining module includes:
Connected domain determination unit for carrying out silhouette markup to the banknote image, and determines the company of the banknote image
Logical domain quantity；
Second dust image determination unit, if being greater than 1 for the connected domain quantity, it is determined that have in the banknote image
Dust image.
Further, the track width determining module includes:
Second difference true ground unit, for determining, third presets adjacent pixel in row pixel in adjacent connected area segmentation region
Second pixel value of point；
Second pixel determination unit, for screening the second maximum value in second pixel value and for the second time most
Big value, and determine second maximum value and the corresponding pixel of second of maximum value；
Second distance determination unit, for determining second maximum value and the corresponding pixel of second of maximum value
Second distance；
Second dust width determination unit, for will be corresponded to described in row pixel in each adjacent connected area segmentation region
The sum of second distance is determined as the dust width of the row pixel of the row pixel；
Second track width determination unit, for the mean value of the dust width to be determined as to the track of the dust image
Width.
Further, described device further include:
Dust width sorting module, for wide in the track that the mean value of the dust width is determined as the dust image
Before degree, the dust width is subjected to numerical ordering；
Correspondingly, the first track width determination unit or the second track width determination unit are specifically used for:
The dust width that predetermined order range is screened in the numerical ordering, the mean value of the dust width after screening is made
For the track width of the dust image.
The embodiment of the present invention includes dust image in determining the image to be detected by obtaining area to be tested image
When, it determines the track width of dust image, if the track width of dust image is greater than predetermined width threshold value, generates dedusting prompting
It instructs and shows, the case where instead of dedusting time is determined by artificial experience in the prior art, solve and artificial determine dedusting
Finance device dedusting caused by time not in time, reduces the problem of paper money recognition accuracy or staff carry out idle work,
The dust deposit state inside automatic checkout equipment is realized, the paper money recognition accuracy of finance device is improved, improves work
Make the working efficiency of personnel.
Detailed description of the invention
Fig. 1 is a kind of flow chart for dust detection method that the embodiment of the present invention one provides；
Fig. 2A is a kind of flow chart of dust detection method provided by Embodiment 2 of the present invention；
Fig. 2 B is a kind of area to be tested image schematic diagram provided by Embodiment 2 of the present invention；
Fig. 3 A is a kind of flow chart for dust detection method that the embodiment of the present invention three provides；
Fig. 3 B is a kind of area to be tested image schematic diagram that the embodiment of the present invention three provides；
Fig. 4 is a kind of structural schematic diagram for dust detection that the embodiment of the present invention four provides.
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining the present invention rather than limiting the invention.It also should be noted that in order to just
Only the parts related to the present invention are shown in description, attached drawing rather than entire infrastructure.
Embodiment one
Fig. 1 is a kind of flow chart for dust detection method that the embodiment of the present invention one provides, and the present embodiment is applicable to certainly
Dynamic the case where carrying out the detection of finance device internal dust, this method can be by a kind of dust detection dresses provided in an embodiment of the present invention
Set to execute, the mode which can be used software and/or hardware is realized, the device can integrate in finance device.Referring to figure
1, this method specifically includes:
S101, area to be tested image is obtained.
In the present embodiment, the influence based on finance device internal dust deposition is mainly reflected in forge or true or paper money identification process
Noise is introduced, paper money recognition error is increased, area to be tested can be the paper money transmission channel region of finance device currency checking module.
In the present embodiment, area to be tested image can be obtained by the conventional images acquisition device of finance device, without increasing
Add new equipment, saves the production cost of finance device.
Optionally, after obtaining image to be detected, mini-value filtering is carried out to the image to be detected.Wherein, minimum value
Filtering can reduce the noise of area to be tested image, reduce due to finance device image acquiring device self performance error, improve
The clarity and accuracy of area to be tested image.
S102, determine whether there is dust image in area to be tested image.
Wherein, dust image refers to the image that the dust inside by finance device is formed, and distinguishes banknote image and finance
Equipment inner structure image.In the present embodiment, finance device is mainly by way of image procossing, if golden to the identification of bank note
Melt inside equipment that there are dust deposits, and dust deposit amount can influence the accuracy of the identification of bank note, then paper money recognition image
In there are dust images.If there is dust image in area to be tested image, it is determined that there are dust deposits for area to be tested, instead
It, if free from dust image in area to be tested image, it is determined that dust deposit or dust deposit is not present very in area to be tested
It is few, it is not enough to influence the identification of bank note.
If having dust image in S103, area to be tested image, it is determined that the track width of dust image.
Wherein, area to be tested image is obtained by way of linear scan, if there are dust in paper money transmission channel,
In the transmission process of bank note, there are dust image in every one-row pixels of the area to be tested image of acquisition, cause entire
The track of dust image is formed in image to be detected.
In the present embodiment, the deposition of dust can determine according to the track width of dust image.
If the track width of S104, dust image is greater than predetermined width threshold value, generates dedusting and remind instruction and show.
Wherein, predetermined width threshold value refers to the maximum that deposition dust is shown in image to be detected inside finance device
Value, illustratively, if the track width of dust image is greater than predetermined width threshold value, the dust deposited will lead to the identification of bank note
Error be greater than default error threshold, influence the normal operation of finance device, predetermined width threshold value for example can be 1.5mm.In advance
If width threshold value can be determined according to history dust testing result, which need to take into account the recognition accuracy and work of bank note
Make the workload of personnel.
In the present embodiment, if the track width of dust image is greater than predetermined width threshold value, generates dedusting and reminds instruction,
The instruction referred to for reminding staff to be dusted finance device is reminded in middle dedusting.Dedusting reminds instruction can be with
It is to be generated by finance device, is sent to finance device server.Server receives dedusting and reminds instruction, is reminded and is instructed according to dedusting
The position signal or encoded information of the finance device of middle carrying are determining to dedusting finance device, will be to dedusting finance device information
Staff is showed, dust removal operation is reminded.
Dedusting is reminded instruction can also be and is shown by the suggestion device that finance device surface is arranged in, exemplary
, it can be and show that instruction is reminded in dedusting by warning light.
Optionally, dedusting grade is determined according to the difference between the track width of dust image and predetermined width threshold value.Show
Example property, if the difference between the track width of dust image and predetermined width threshold value is bigger, it is heavy inside equipment to show to enter
Long-pending dust is more, bigger on paper money recognition influence, and dedusting higher grade；On the contrary, if the track width of dust image and default
Difference between width threshold value is smaller, then shows that the dust for entering deposition inside equipment is fewer, smaller on paper money recognition influence, removes
Dirt lower grade.In the present embodiment, the pressing degree and sequencing of finance device dedusting can determine by dedusting grade, improve
The working efficiency of staff.
The technical solution of the present embodiment by obtaining area to be tested image, and in determining the image to be detected includes
When dust image, the track width of dust image is determined, if the track width of dust image is greater than predetermined width threshold value, generate
Dedusting is reminded to instruct and simultaneously be shown, the case where instead of dedusting time is determined by artificial experience in the prior art, is solved manually
It determines that finance device dedusting is not in time caused by dedusting time, reduces paper money recognition accuracy or staff carries out idle work
The problem of, the dust deposit state inside automatic checkout equipment is realized, the paper money recognition accuracy of finance device is improved, mentions
The high working efficiency of staff.
Embodiment two
Fig. 2A is a kind of flow chart of dust detection method provided by Embodiment 2 of the present invention, in the base of above-described embodiment one
It is further refined on plinth.Correspondingly, A referring to fig. 2, this method comprises:
S201, the gray level image for obtaining transmission channel.
Wherein, transmission channel refers to the transmission channel of bank note inside finance device, and in the present embodiment, transmission channel can be with
It is to be located at finance device currency checking module.
S202, according to the relative position information of idler wheel and transmission channel, determine the idler wheel picture position of gray level image, idler wheel
In transmission channel.
In the present embodiment, idler wheel is the component part of transmission channel, and the movement of idler wheel drives the transmission of bank note.The present embodiment
In, the dust in transmission channel, which is generally concentrated, to be deposited on idler wheel, and influence of the dust at idler wheel to paper money recognition is maximum.
Idler wheel and the relative position information of transmission channel are fixed, and the relative position information can be directly read.Illustratively, it passes
Defeated channel width is Lcm, and idler wheel is arranged at the M1cm-M2cm of transmission channel, then can be true in the dust image of transmission channel
Determine the position of idler wheel image, such as idler wheel picture position is (X1, X2)={ DPI* (M1/L), DPI* (M2/L) }, wherein X1 and
X2 be idler wheel image edge location, DPI be unit distance comprising pixel quantity, DPI can be according to finance device internal image
Acquisition device determines.Such as DPI can be 100-200.
S203, area to be tested image is intercepted in the gray level image of transmission channel according to idler wheel picture position, it is to be detected
Area image includes background image and banknote image, and banknote image is located inside background image region.
In the present embodiment, area to be tested image is intercepted according to scroll wheel positions, dust detection range is reduced, improves ash
The specific aim of dirt detection, improves the efficiency of dust detection.
Illustratively, B, Fig. 2 B are a kind of area to be tested image schematic diagrames provided by Embodiment 2 of the present invention referring to fig. 2.
Area to be tested image includes banknote image 210 and background image 220 in Fig. 2 B, and banknote image 210 is located at background image region
Inside 220.
In the present embodiment, every one banknote passes through transmission channel with Fixed Time Interval, and image acquiring device is solid according to this
Interval acquiring of fixing time includes the gray level image of the transmission channel of banknote image.
S204, determine in background image the accumulative difference of the pixel value of the first default row pixel and.
Wherein, the accumulative difference of pixel value and the pixel value absolute value for referring to each neighbor pixel in row pixel
And value, the pixel value for being illustratively encoded to the row pixel of n add up difference and can be determined by following formula:
Wherein, Rdiffsum (n) is that the pixel value for the row pixel for being encoded to n adds up difference and i is pixel in row pixel
Coding, w be row pixel in include pixel quantity, PniFor the pixel value of line n pixel ith pixel point.
In the present embodiment, the default row pixel of the first of background image is the present count in the predeterminated position selection of background image
The row pixel of amount, for example, background image the first default row pixel can be background image borderline region preset quantity row
Pixel.Illustratively, B, background image include a quarter each up and down region of area to be tested image referring to fig. 2, if such as
The row pixel quantity of area to be tested image is 1024, is encoded to all row pixels, then background image includes to be encoded to
The row pixel of 1-256 and 768-1024.
The default row pixel of the first of background image can be the institute in a quarter each up and down region of area to be tested image
There is row pixel, be also possible to the row pixel of preset quantity in a quarter each up and down region according to area to be tested image,
Middle preset quantity can be to be determined according to history dust testing result.
Optionally, step S204 may also is that the pixel value of the first default row pixel in determining area to be tested image is tired
Count difference and.
If S205, adding up difference and being greater than default difference and threshold value, it is determined that there is dust image in background image.
Illustratively, background image color is approximately black in B referring to fig. 2, Fig. 2 B, if there is no ashes in background image
Dirt, then the pixel value of each neighbor pixel is smaller, the accumulative difference of row pixel and smaller；If there is ash in background image
Dirt, then the pixel value of dust pixel and background pixel point is larger, the accumulative difference of row pixel and larger.Default difference and
Threshold value is according to pixel quantity determines in row pixel in image to be detected, and pixel quantity is bigger in row pixel, and it is poor to preset
Divide and threshold value is bigger, conversely, pixel quantity is smaller in row pixel, presets difference and threshold value is smaller.
Illustratively, B referring to fig. 2 includes dust image 230 in background image 220, and determines the ash of dust image 230
Dirt type is light dust.Wherein light dust is mainly since the paper scrap that bank note is fallen in transmission process deposits on idler wheel
It is formed, since the dust color that paper scrap is formed is shallower, it is biggish to form brightness in the background image of area to be tested image
Dust image.
In the present embodiment, detects the accumulative difference of each row pixel and whether be greater than default difference and threshold value, can be to each
It the accumulative difference of row pixel and is detected respectively, the mean value for being also possible to detect the accumulative difference sum of each row pixel is examined
It surveys.If the accumulative difference of each row pixel and being greater than default difference and threshold value, it is determined that have dust image in background image, and continue
Execute step S206；If not, it is determined that dust image is not present in background image, and returns to step S201, re-starts
Dust detection.
S206, the first pixel value for obtaining each neighbor pixel in the second default row pixel of image to be detected.
Wherein, the second default row pixel refers to the row pixel in background image comprising dust image.In the present embodiment, the
Two default row pixels can be identical as the first default row pixel, be also possible to different from the first row pixel.Illustratively, second
Default row pixel can be including being encoded to preceding 1/4 row pixel in area to be tested image and/or being encoded to rear 1/4 row picture
Element, if such as area to be tested image row pixel quantity be 1024, all row pixels are encoded, then the second default row picture
Element can be the row pixel including being encoded to 1-256 and/or 768-1024.
In the present embodiment, the first pixel value of each neighbor pixel can be the pixel value of each neighbor pixel
Absolute value.
S207, the first maximum value in the first pixel value of screening and first time maximum difference, and determine that first is maximum
It is worth pixel corresponding with first time maximum value.
First pixel value in each row pixel is subjected to descending sequence, first difference and second in sequence
Difference is respectively the first maximum value and first time maximum difference.
S208, the first maximum value in each row pixel and the first distance of the corresponding pixel of first time maximum value are determined as
The dust width of each row pixel.
In the present embodiment, B referring to fig. 2, the gray value of each pixel is smaller in background image 220, and neighbor pixel
Pixel value is smaller.Dust image 230 is light dust, and the pixel value for the pixel for including in dust image is larger, with back
The difference of the pixel value of pixel in scape image is larger, and the pixel value of adjacent dust pixel is smaller.Pass through screening
The first maximum value and first time maximum difference in first pixel value, it may be determined that pixel value in pixel of being expert at is maximum
Above-mentioned two pixel, is determined as the boundary pixel point of dust image in row pixel by two pixels, and by dust image
The distance between boundary pixel point is determined as the dust width in the row pixel.
S209, the track width that the mean value of dust width is determined as to dust image.
In the present embodiment, every one-row pixels in the second default row pixel can determine a dust width, in order to avoid
There is error or mistake in part dust width, the track that the average value of each dust width is determined as dust image is wide
Degree improves the accuracy of dust image path width, improves dust detection accuracy.
If the track width of S210, dust image is greater than predetermined width threshold value, generates dedusting and remind instruction and show.
It should be noted that step S201-S210 is illustratively formed one embodiment by the present invention executes a kind of bank note
Detection method, but a kind of example only of the invention, in other embodiments of the invention, can by step S201-203,
S102-S104 forms a new embodiment, or step S201-205, S103-S104 can also be formed one embodiment,
A kind of dust detection method can be performed in above-described embodiment.
The technical solution of the present embodiment, by the first default row pixel for determining the background image in area to be tested image
The accumulative difference of pixel value and, determine whether area to be tested image includes dust according to accumulative difference and default difference and threshold value
Image, if in area to be tested image including dust image, according to the dust width for determining the second default row pixel, according to each
The mean value of dust width determines the track width of dust image, and determines whether that generating dedusting reminds instruction, solves existing skill
The problem of in art by manually determining finance device dedusting time, the dust deposit state inside automatic checkout equipment is realized, is mentioned
The high paper money recognition accuracy of finance device, improves the working efficiency of staff.
On the basis of the above embodiments, may include: before step S209
Each dust width is subjected to numerical ordering；
Correspondingly, step S209 can be the dust width for screening predetermined order range in numerical ordering, after screening
Dust width track width of the mean value as dust image.
In the present embodiment, in the transmission process of bank note, tilt of paper money be may cause in the second default row pixel comprising portion
Divide banknote image, the track width accuracy in computation of dust image is caused to reduce.
The dust width of each row pixel is subjected to numerical ordering, such as is ranked up from small to large or from big to small, is sieved
Select the predetermined order range dust width in numerical ordering as effective dust width, wherein predetermined order range can be according to history
Dust testing result determines, such as predetermined order range can be 40%-60% range in numerical ordering.By effective ash of screening
Track width of the mean value of dirt width as dust image eliminates error number caused by as calculating error or image error
According to improving the accuracy in computation of the track width of dust image, improve dust detection accuracy.
Embodiment three
Fig. 3 A is a kind of flow chart for dust detection method that the embodiment of the present invention three provides, on the basis of above-described embodiment
On further refined.
S301, the gray level image for obtaining transmission channel.
S302, according to the relative position information of idler wheel and transmission channel, determine the idler wheel picture position of gray level image, idler wheel
In transmission channel.
S303, area to be tested image is intercepted in the gray level image of transmission channel according to idler wheel picture position, it is to be detected
Area image includes background image and banknote image, and banknote image is located inside background image region.
S304, silhouette markup is carried out to banknote image, and determine the connected domain quantity of banknote image.
Wherein, silhouette markup refers to the pixel by traversing banknote image, marks Confined outline curve, closed outline
Region in curve is a connected domain, and the quantity of Confined outline curve is identical as connected domain quantity.
If S305, connected domain quantity are greater than 1, it is determined that there is dust image in banknote image.
Illustratively, if connected domain quantity is 1 in banknote image, it is determined that only comprising by bank note boundary group in banknote image
At a Confined outline curve, free from dust image in banknote image；If connected domain quantity is greater than 1 in banknote image, it is determined that
Include at least two Confined outline curves being made of bank note boundary and dust image in banknote image, there is ash in banknote image
Dirt image.Referring to Fig. 3 B, Fig. 3 B is a kind of area to be tested image schematic diagram of the embodiment of the present invention but offer.In Fig. 3 B, to
Detection zone image includes background image 320, banknote image 310 and dust image 330, dust image 330 and banknote image 310
Boundary formed two Confined outline curves, banknote image 310 is divided into two connected domains.
In the present embodiment, if banknote image includes dust image, the dust image can determine for dark dust, wherein deep
During color dust mainly passes through paper currency transmission, the foreign matters such as soil, spot for carrying on bank note are attached to be formed on idler wheel.
S306, the second margin of image element for determining neighbor pixel in the default row pixel of third in adjacent connected area segmentation region
Value.
In the present embodiment, adjacent connected area segmentation region refers to the adjacent public contour curve region of connected domain, shows
Example property, referring to Fig. 3 B, in Fig. 3 B, region 340 is the cut zone of two connected domains in banknote image 310.If banknote image
The quantity of middle connected domain is N, it is determined that comprising a adjacent connected area segmentation regions N-1 dust image and N-1 in banknote image,
Wherein N is the positive integer more than or equal to 1.
Wherein, the default row pixel of third refers to the row pixel in banknote image comprising dust image.Third presets row picture
Element can be including row pixels all in banknote image, and it is pre- as third to be also possible to the selector branch pixel in banknote image
If row pixel.
Illustratively, third presets the center row pixel and center row picture that can be through area to be tested of row pixel
The row pixel and/or the row total number of pixels 1/4 after center row pixel coder of row total number of pixels 1/4 before element coding
Row pixel, if such as area to be tested image row pixel quantity be 1024, all row pixels are encoded, then third is pre-
If row pixel includes the row pixel for being encoded to 257-512 and/or being encoded to 512-768.
The second maximum value and second of maximum value in S307, the second pixel value of screening, and determine the second maximum value
Pixel corresponding with second of maximum value.
In the present embodiment, if adjacent connected area segmentation region is greater than 1, determined in each adjacent connected area segmentation region
Second pixel value of the neighbor pixel of each row pixel, and the second pixel is screened in each adjacent connected area segmentation region
The second maximum value and second of maximum value and corresponding pixel in value difference value.
S308, the second distance for determining the second maximum value and the corresponding pixel of second of maximum value.
S309, the ash that the sum of second distance that row pixel is corresponded in each adjacent connected area segmentation region is determined as to row pixel
Dirt width；
In the present embodiment, if adjacent connected area segmentation region quantity is 1, second in the adjacent connected area segmentation region
Distance is determined as the dust width of row pixel, if adjacent connected area segmentation region quantity is greater than 1, shows that dust amount of images is greater than
1, then the sum of the second distance that row pixel is corresponded in each adjacent connected area segmentation region is determined as to the dust width of row pixel,
In, corresponding row pixel refers to row pixel identical in each adjacent connected area segmentation region interior coding.It avoids and works as banknote image
In there are multiple dust images, and single dust image dust width it is smaller when, cannot correctly detect the precipitation number of dust
Amount leads to the situation of dust detection inaccuracy, improves dust detection accuracy.
S310, the track width that the mean value of dust width is determined as to dust image.
If the track width of S311, dust image is greater than predetermined width threshold value, generates dedusting and remind instruction and show.
It should be understood that step S301-S311 is illustratively formed one embodiment by the present invention executes a kind of bank note
Detection method, but a kind of example only of the invention, in other embodiments of the invention, can by step S201-205,
S103-S104 forms one embodiment, and a kind of dust detection method can be performed in above-described embodiment.
It should be noted that the technical solution of the present embodiment, synchronous with the technical solution in embodiment two can execute, if to
There are dust images for any portion in background image and banknote image in detection zone image, and the track width of dust image is big
In predetermined width threshold value, then generates dedusting and remind instruction and show；If background image and banknote image in area to be tested image
In there is no dust image or existing dust image track width be less than predetermined width threshold value, then return under acquisition
One area to be tested image, circulation carry out dust detection.
The technical solution of the present embodiment carries out silhouette markup by the banknote image to area to be tested image, determines paper
Connected domain quantity in coin image determines in banknote image comprising dust image when connected domain quantity is greater than 1, determines bank note
Third presets the dust width of row pixel in image, and the track width of dust image is determined according to the mean value of each dust width, and
Determine whether that generating dedusting reminds instruction, solves the problems, such as in the prior art by manually determining finance device dedusting time, it is real
The dust deposit state inside automatic checkout equipment is showed, has improved the paper money recognition accuracy of finance device, improve work
The working efficiency of personnel.
On the basis of the above embodiments, may include: before step S310
Each dust width is subjected to numerical ordering；
Correspondingly, step S310 can be the dust width for screening predetermined order range in numerical ordering, after screening
Dust width track width of the mean value as dust image.
In the present embodiment, the predetermined order range of banknote image and the predetermined order range of background image can be identical
, be also possible to it is different, predetermined order range can be according to carry out numerical ordering dust width quantity determine.It will screening
Effective dust width track width of the mean value as dust image, caused by eliminating as calculating error, image error
Wrong data improves the accuracy in computation of the track width of dust image, improves dust detection accuracy.
Example IV
Fig. 4 is a kind of structural schematic diagram for dust detection that the embodiment of the present invention four provides, which specifically includes:
Image collection module 410, for obtaining area to be tested image；
Dust image determining module 420, for determining whether there is dust image in area to be tested image；
Track width determining module 430, if for having dust image in area to be tested image, it is determined that dust image
Track width；
Directive generation module 440 is reminded in dedusting, if the track width for dust image is greater than predetermined width threshold value, is given birth to
It is reminded at dedusting and instructs and show.
Optionally, image collection module 410 includes:
Transmission channel image acquisition unit, for obtaining the gray level image of transmission channel；
Idler wheel picture position determination unit determines grayscale image for the relative position information according to idler wheel and transmission channel
The idler wheel picture position of picture, idler wheel are located in transmission channel；
Image interception unit, for intercepting area to be tested in the gray level image of transmission channel according to idler wheel picture position
Image, area to be tested image include background image and banknote image, and banknote image is located inside background image region.
Optionally, dust image determining module 420 includes:
Accumulative difference and determination unit, for determining the accumulative difference of the pixel value of the first default row pixel in background image
With；
First dust image determination unit, if for accumulative difference and being greater than default difference and threshold value, it is determined that Background
There is dust image as in.
Optionally, track width determining module 430 includes:
First difference true ground unit, the of each neighbor pixel in the second default row pixel for obtaining image to be detected
One pixel value；
First pixel determination unit, it is maximum poor for screening the first maximum value in the first pixel value and first time
Value, and determine the first maximum value and the corresponding pixel of first time maximum value；
First dust width determination unit is used for the first maximum value in each row pixel and the corresponding picture of first time maximum value
The first distance of vegetarian refreshments is determined as the dust width of each row pixel；
First track width determination unit, for the mean value of dust width to be determined as to the track width of dust image.
Optionally, dust image determining module 420 includes:
Connected domain determination unit for carrying out silhouette markup to banknote image, and determines the connected domain quantity of banknote image；
Second dust image determination unit, if being greater than 1 for connected domain quantity, it is determined that there is dust figure in banknote image
Picture.
Optionally, track width determining module 430 includes:
Second difference true ground unit, for determining, third presets adjacent pixel in row pixel in adjacent connected area segmentation region
Second pixel value of point；
Second pixel determination unit, it is maximum for screening the second maximum value in the second pixel value and second
Value, and determine the second maximum value and the corresponding pixel of second of maximum value；
Second distance determination unit, for determine the second of the second maximum value and the corresponding pixel of second of maximum value away from
From；
Second dust width determination unit, for the second distance of row pixel will to be corresponded in each adjacent connected area segmentation region
The sum of be determined as the dust width of row pixel；
Second track width determination unit, for the mean value of dust width to be determined as to the track width of dust image.
Optionally, device further include:
Dust width sorting module, for before the mean value of dust width is determined as the track width of dust image,
Dust width is subjected to numerical ordering；
Correspondingly, the first track width determination unit or the second track width determination unit are specifically used for:
The dust width that predetermined order range is screened in numerical ordering, using the mean value of the dust width after screening as ash
The track width of dirt image.
Dust provided by any embodiment of the invention can be performed in a kind of dust detection provided in an embodiment of the present invention
Detection method has and executes the corresponding functional module of dust detection method and beneficial effect.
Note that the above is only a better embodiment of the present invention and the applied technical principle.It will be appreciated by those skilled in the art that
The invention is not limited to the specific embodiments described herein, be able to carry out for a person skilled in the art it is various it is apparent variation,
It readjusts and substitutes without departing from protection scope of the present invention.Therefore, although being carried out by above embodiments to the present invention
It is described in further detail, but the present invention is not limited to the above embodiments only, without departing from the inventive concept, also
It may include more other equivalent embodiments, and the scope of the invention is determined by the scope of the appended claims.
Claims (10)
1. a kind of dust detection method characterized by comprising
Obtain area to be tested image；
Determine in the area to be tested image whether there is dust image；
If so, determining the track width of the dust image；
If the track width of the dust image is greater than predetermined width threshold value, generates dedusting and remind instruction and show.
2. the method according to claim 1, wherein obtaining area to be tested image, comprising:
Obtain the gray level image of transmission channel；
According to the relative position information of idler wheel and the transmission channel, the idler wheel picture position of the gray level image is determined, it is described
Idler wheel is located in the transmission channel；
The area to be tested image is intercepted in the gray level image of the transmission channel according to the idler wheel picture position, it is described
Area to be tested image includes background image and banknote image, and the banknote image is located inside the background image region.
3. according to the method described in claim 2, it is characterized in that, determining in the area to be tested image whether there is dust figure
Picture, comprising:
Determine in the background image the accumulative difference of the pixel value of the first default row pixel and；
If the accumulative difference and being greater than default difference and threshold value, it is determined that there is dust image in the background image.
4. according to the method described in claim 3, it is characterized in that, determining the track width of the dust image, comprising:
Obtain the first pixel value of each neighbor pixel in the second default row pixel of described image to be detected；
The first maximum value and first time maximum difference in first pixel value are screened, and determines first maximum value
Pixel corresponding with the first time maximum value；
The first distance of first maximum value described in each row pixel and the corresponding pixel of the first time maximum value is determined as
The dust width of each row pixel；
The mean value of the dust width is determined as to the track width of the dust image.
5. according to the method described in claim 2, it is characterized in that, determining in the area to be tested image whether there is dust figure
Picture, comprising:
Silhouette markup is carried out to the banknote image, and determines the connected domain quantity of the banknote image；
If the connected domain quantity is greater than 1, it is determined that there is dust image in the banknote image.
6. according to the method described in claim 5, it is characterized in that, determining the track width of the dust image, comprising:
Determine the second pixel value of neighbor pixel in the default row pixel of third in adjacent connected area segmentation region；
Screen the second maximum value and second of maximum value in second pixel value, and determine second maximum value and
The corresponding pixel of second of maximum value；
Determine the second distance of second maximum value and the corresponding pixel of second of maximum value；
The sum of the second distance that row pixel is corresponded in each adjacent connected area segmentation region is determined as the row pixel
Dust width；
The mean value of the dust width is determined as to the track width of the dust image.
7. the method according to claim 4 or 6, which is characterized in that described in being determined as the mean value of the dust width
Before the track width of dust image, comprising:
The dust width is subjected to numerical ordering；
Correspondingly, the mean value of the dust width to be determined as to the track width of the dust image, comprising:
The dust width that predetermined order range is screened in the numerical ordering, using the mean value of the dust width after screening as institute
State the track width of dust image.
8. a kind of dust detection characterized by comprising
Image collection module, for obtaining area to be tested image；
Dust image determining module, for determining in the area to be tested image whether there is dust image；
Track width determining module, if for having dust image in the area to be tested image, it is determined that the dust image
Track width；
Directive generation module is reminded in dedusting, if the track width for the dust image is greater than predetermined width threshold value, is generated
Dedusting, which is reminded, to be instructed and shows.
9. device according to claim 8, which is characterized in that described image obtains module and includes:
Transmission channel image acquisition unit, for obtaining the gray level image of transmission channel；
Idler wheel picture position determination unit determines the ash for the relative position information according to idler wheel and the transmission channel
The idler wheel picture position of image is spent, the idler wheel is located in the transmission channel；
Image interception unit, for intercepted in the gray level image of the transmission channel according to the idler wheel picture position it is described to
Detection zone image, the area to be tested image include background image and banknote image, and the banknote image is located at the back
Inside scape image-region.
10. device according to claim 9, which is characterized in that the dust image determining module includes:
Accumulative difference and determination unit, for determining the accumulative difference of the pixel value of the first default row pixel in the background image
With；
First dust image determination unit, if for the accumulative difference and being greater than default difference and threshold value, it is determined that the back
There is dust image in scape image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710347931.2A CN108961200A (en) | 2017-05-17 | 2017-05-17 | A kind of dust detection method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710347931.2A CN108961200A (en) | 2017-05-17 | 2017-05-17 | A kind of dust detection method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108961200A true CN108961200A (en) | 2018-12-07 |
Family
ID=64462509
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710347931.2A Pending CN108961200A (en) | 2017-05-17 | 2017-05-17 | A kind of dust detection method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108961200A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109459354A (en) * | 2019-01-20 | 2019-03-12 | 华东交通大学 | A kind of workshop dust fall amount automatic detection device and method based on machine vision |
| CN109990711A (en) * | 2019-04-25 | 2019-07-09 | 湘潭大学 | A kind of appearance quality detection method of punched nickel-plated steel band |
| CN110310730A (en) * | 2019-07-02 | 2019-10-08 | 韩瑞霞 | Infect department's real-time data analysis device and method |
| CN110755966A (en) * | 2019-11-07 | 2020-02-07 | 黄雨果 | Intelligent tire water spray dust removal control platform |
| CN111257335A (en) * | 2020-01-09 | 2020-06-09 | Oppo（重庆）智能科技有限公司 | Method for detecting dust points in electronic equipment |
| CN111915792A (en) * | 2020-05-19 | 2020-11-10 | 武汉卓目科技有限公司 | Method and device for identifying zebra crossing image-text |
| CN115049662A (en) * | 2022-08-16 | 2022-09-13 | 山东大拇指喷雾设备有限公司 | Sprayer flow control method based on image processing |
| CN115496759A (en) * | 2022-11-17 | 2022-12-20 | 歌尔股份有限公司 | Dust detection method and device and storage medium |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE69814495D1 (en) * | 1997-03-06 | 2003-06-18 | Xerox Corp | Method and device for automatic edge detection of a document using a scanner |
| CN101494743A (en) * | 2008-01-23 | 2009-07-29 | Hoya株式会社 | Dust detection system and digital camera |
| CN103279735A (en) * | 2013-04-24 | 2013-09-04 | 广州广电运通金融电子股份有限公司 | Dust stratification detection method and system for financial bill identification module |
| CN103357672A (en) * | 2012-03-30 | 2013-10-23 | 鞍钢股份有限公司 | Online strip steel boundary detection method |
| CN104200566A (en) * | 2014-09-11 | 2014-12-10 | 广州广电运通金融电子股份有限公司 | Banknote recognition method based on sorter gradation accumulation and sorter |
| CN105225335A (en) * | 2015-10-28 | 2016-01-06 | 深圳怡化电脑股份有限公司 | A kind of stained recognition methods of paper money number and system |
| CN105427448A (en) * | 2015-10-29 | 2016-03-23 | 深圳怡化电脑股份有限公司 | Method and device of obtaining banknote security thread boundary |
| CN106127930A (en) * | 2016-06-17 | 2016-11-16 | 深圳怡化电脑股份有限公司 | A kind of method and system detecting banknote security line integrity |
| CN106408747A (en) * | 2016-09-06 | 2017-02-15 | 深圳怡化电脑股份有限公司 | Image recognition method and apparatus |
| CN106469482A (en) * | 2016-09-05 | 2017-03-01 | 深圳怡化电脑股份有限公司 | Bank note crosses the border detection method, device and Banknote deposting/drawing out equipment |
-
2017
- 2017-05-17 CN CN201710347931.2A patent/CN108961200A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE69814495D1 (en) * | 1997-03-06 | 2003-06-18 | Xerox Corp | Method and device for automatic edge detection of a document using a scanner |
| CN101494743A (en) * | 2008-01-23 | 2009-07-29 | Hoya株式会社 | Dust detection system and digital camera |
| CN103357672A (en) * | 2012-03-30 | 2013-10-23 | 鞍钢股份有限公司 | Online strip steel boundary detection method |
| CN103279735A (en) * | 2013-04-24 | 2013-09-04 | 广州广电运通金融电子股份有限公司 | Dust stratification detection method and system for financial bill identification module |
| CN104200566A (en) * | 2014-09-11 | 2014-12-10 | 广州广电运通金融电子股份有限公司 | Banknote recognition method based on sorter gradation accumulation and sorter |
| CN105225335A (en) * | 2015-10-28 | 2016-01-06 | 深圳怡化电脑股份有限公司 | A kind of stained recognition methods of paper money number and system |
| CN105427448A (en) * | 2015-10-29 | 2016-03-23 | 深圳怡化电脑股份有限公司 | Method and device of obtaining banknote security thread boundary |
| CN106127930A (en) * | 2016-06-17 | 2016-11-16 | 深圳怡化电脑股份有限公司 | A kind of method and system detecting banknote security line integrity |
| CN106469482A (en) * | 2016-09-05 | 2017-03-01 | 深圳怡化电脑股份有限公司 | Bank note crosses the border detection method, device and Banknote deposting/drawing out equipment |
| CN106408747A (en) * | 2016-09-06 | 2017-02-15 | 深圳怡化电脑股份有限公司 | Image recognition method and apparatus |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109459354A (en) * | 2019-01-20 | 2019-03-12 | 华东交通大学 | A kind of workshop dust fall amount automatic detection device and method based on machine vision |
| CN109990711A (en) * | 2019-04-25 | 2019-07-09 | 湘潭大学 | A kind of appearance quality detection method of punched nickel-plated steel band |
| CN109990711B (en) * | 2019-04-25 | 2021-08-31 | 湘潭大学 | Appearance quality detection method for punched nickel-plated steel strip |
| CN110310730A (en) * | 2019-07-02 | 2019-10-08 | 韩瑞霞 | Infect department's real-time data analysis device and method |
| CN110755966A (en) * | 2019-11-07 | 2020-02-07 | 黄雨果 | Intelligent tire water spray dust removal control platform |
| CN110755966B (en) * | 2019-11-07 | 2021-06-29 | 黄雨果 | Intelligent tire water spray dust removal control platform |
| CN111257335A (en) * | 2020-01-09 | 2020-06-09 | Oppo（重庆）智能科技有限公司 | Method for detecting dust points in electronic equipment |
| CN111915792A (en) * | 2020-05-19 | 2020-11-10 | 武汉卓目科技有限公司 | Method and device for identifying zebra crossing image-text |
| CN111915792B (en) * | 2020-05-19 | 2022-06-07 | 武汉卓目科技有限公司 | Method and device for identifying zebra crossing image-text |
| CN115049662A (en) * | 2022-08-16 | 2022-09-13 | 山东大拇指喷雾设备有限公司 | Sprayer flow control method based on image processing |
| CN115049662B (en) * | 2022-08-16 | 2022-11-08 | 山东大拇指喷雾设备有限公司 | Sprayer flow control method based on image processing |
| CN115496759A (en) * | 2022-11-17 | 2022-12-20 | 歌尔股份有限公司 | Dust detection method and device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108961200A (en) | A kind of dust detection method and device | |
| CN104794491B (en) | Based on the fuzzy clustering Surface Defects in Steel Plate detection method presorted | |
| CN104091388B (en) | A kind of paper money discrimination method based on magnetic image and device | |
| CN101923741B (en) | Paper currency number identification method based on currency detector | |
| CN102043959B (en) | License plate character segmentation method | |
| CN106780966B (en) | A kind of paper money discrimination method and device | |
| CN101433075A (en) | Generating a bitonal image from a scanned colour image | |
| WO2015120193A1 (en) | Defect detection and classification based on attributes determined from a standard reference image | |
| US7209582B2 (en) | Coin discriminating method and apparatus | |
| CN106780486A (en) | A kind of Surface Defects in Steel Plate image extraction method | |
| CN104766079B (en) | A kind of remote method for detecting infrared puniness target | |
| US20090034810A1 (en) | Abnormal tissue pattern detection apparatus, method and program | |
| WO2016037523A1 (en) | Banknote recognition method based on sorter dust accumulation and sorter | |
| CN109685760A (en) | A kind of SLM powder bed powder laying image convex closure depression defect detection method based on MATLAB | |
| CN103606221A (en) | Fault automatic diagnostic method of counter and device | |
| CN103093185A (en) | Character recognition device, image processing device and methods thereof | |
| CN106780464A (en) | A kind of fabric defect detection method based on improvement Threshold segmentation | |
| CN111189854B (en) | Defect layering detection method of automatic glass cover plate detection system | |
| KR20150108118A (en) | Remote automatic metering system based image recognition | |
| CN106204616B (en) | Method and device for identifying currency value of Iran paper money | |
| CN109030499A (en) | A kind of device and method preventing defect counts repeat count suitable for target defect continuous on-line detection | |
| CN116596921A (en) | Method and system for sorting incinerator slag | |
| CN201804365U (en) | Coin image recognition device | |
| CN113205633B (en) | Method and device for detecting red and blue fibers of bill, electronic equipment and storage medium | |
| CN114937003A (en) | Multi-type defect detection system and method for glass panel |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181207 |