CN107194887A - A kind of image deblurring method and system applied to car plate detection - Google Patents
A kind of image deblurring method and system applied to car plate detection Download PDFInfo
- Publication number
- CN107194887A CN107194887A CN201710316939.2A CN201710316939A CN107194887A CN 107194887 A CN107194887 A CN 107194887A CN 201710316939 A CN201710316939 A CN 201710316939A CN 107194887 A CN107194887 A CN 107194887A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- image
- deblurring
- gray level
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 29
- 238000000034 method Methods 0.000 title claims abstract description 26
- 238000006243 chemical reaction Methods 0.000 claims abstract description 40
- 230000009467 reduction Effects 0.000 claims abstract description 35
- 238000000527 sonication Methods 0.000 claims abstract description 20
- 230000005540 biological transmission Effects 0.000 claims description 12
- 230000008859 change Effects 0.000 claims description 7
- 241000208340 Araliaceae Species 0.000 claims 2
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims 2
- 235000003140 Panax quinquefolius Nutrition 0.000 claims 2
- 235000008434 ginseng Nutrition 0.000 claims 2
- 230000000694 effects Effects 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 12
- 238000005516 engineering process Methods 0.000 description 6
- 230000010354 integration Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 239000003086 colorant Substances 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000007935 neutral effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/73—Deblurring; Sharpening
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of image deblurring method and system applied to car plate detection, vehicle high-speed movement is mainly solved, relative movement causes the fuzzy of license plate image between automobile and video camera.This method includes：Obtain coloured image；Gradation conversion is carried out to the coloured image, gray level image is obtained；Noise reduction sonication is carried out to the gray level image, the gray level image of noise reducing is obtained；Deblurring processing is carried out to the gray level image of the noise reducing, the gray level image of noise reducing deblurring is obtained；Binary conversion treatment is carried out to the gray level image of the noise reducing deblurring, bianry image is obtained.By implementing the embodiment of the present invention, the binary image of gradation conversion, noise reduction sonication, deblurring processing and the high accuracy of binary conversion treatment acquisition is carried out successively to the image of car plate detection, with more preferable effect.
Description
Technical field
The present invention relates to technical field of image processing, more particularly to a kind of image deblurring method applied to car plate detection
And system.
Background technology
With expanding economy, the continuous improvement of living standards of the people, increasing automobile, which is put into, to be used, and is increased
Related department further successfully shoots automotive license plate for video camera and proposes higher want to the management difficulty of automobile
Ask.
The characteristics of there was only unique car plate according to an automobile, many companies and related scientific research mechanism devise license plate knowledge
Other technology, mainly including steps such as License Plate, License Plate Segmentation, character recognition.According to spies such as the shape of car plate, edge, colors
Levy, and the technology such as combining form, neutral net carries out License Plate, then using license plate area texture than more rich spy
The methods such as point, car plate Gray Projection carry out License Plate Segmentation, then the image matrix split are mapped again, template matches etc.
Method carries out character recognition.Although these License Plate Recognition Technologies can recognize that car plate under certain condition, in automobile
In the case of high-speed mobile, the fuzzy license plate image still None- identified that these technologies are shot to video camera.
Because the translational speed of automobile is very fast, it is necessary to decoded within the time short enough to image, and due to car
High-speed mobile, the relative movement between automobile and video camera causes the fuzzy of license plate image, causes the license plate shot
Accuracy when image is handled.
The content of the invention
It is an object of the invention to overcome the deficiencies in the prior art, the invention provides a kind of figure applied to car plate detection
As deblurring method system, the binaryzation of the image high accuracy of car plate detection can be effectively improved.
In order to solve the above-mentioned technical problem, a kind of image deblurring method applied to car plate detection of the embodiment of the present invention,
Methods described includes：
Obtain coloured image；
Gradation conversion is carried out to the coloured image, gray level image is obtained；
Noise reduction sonication is carried out to the gray level image, the gray level image of noise reducing is obtained；
Deblurring processing is carried out to the gray level image of the noise reducing, the gray level image of noise reducing deblurring is obtained；
Binary conversion treatment is carried out to the gray level image of the noise reducing deblurring, bianry image is obtained.
Preferably, it is described that gradation conversion is carried out to the coloured image, including the coloured image is turned using brightness value
The mode of changing carries out gradation conversion processing, and the brightness value conversion formula is as follows：
I=0.3 × R+0.59 × G+0.11 × B
Wherein, I represents the brightness value of gray level image, and R represents red, and G represents green, and B represents blueness.
Preferably, it is described that gray level image progress noise reduction sonication is included：
Noise reduction sonication is carried out to the gray level image using weighting total variation algorithm, the formula of total variation algorithm is weighted such as
Under：
Or：
Wherein, i, j represent that 1,2,3..., y represents signal respectively；
Further, a signal x with random noise is givenn, one is obtained close to xnSignal ynWith smaller
Total variation, can be weighed, E (x, y) algorithmic formula is as follows with fidelity bound term E (x, y)：
Wherein, n represents 1,2,3..., and the fidelity bound term E (x, y) combines weighting total variation algorithm, solves minimum
Value, formula is as follows：
E (x, y)+λ V (y)
Wherein, λ represents weight.
Preferably, the gray level image to the noise reducing carries out deblurring processing, including：
To the fuzzy core φ of a transfer functionbEstimated, obtain the fuzzy core φ of point transfer functionb；
According to described transfer function fuzzy core φb, deblurring processing is carried out to the gray level image of the noise reducing；
The fuzzy core φ of described pair of point transfer functionbCarry out algorithm for estimating as follows：
Wherein, λ1＞ 0 represents parameter, u1Represent the image after noise reducing, φbPoint transmission fuzzy core is represented, C1 represents area
Domain, L represents the length of side, I1Represent original image；
Further, on the basis of variational problem is strict convex surface, the fuzzy core φ of described transfer functionbIt is unique minimum
Value point is exactly the solution of Euler-Lagrange equation：
-Δφb+λ1(φ**I1-u1)*I1=0
Wherein, Δ φbRepresent that electrical transmission obscures the difference of sum；
According to described transfer function fuzzy core φb, deblurring processing, algorithm are carried out to the gray level image of the noise reducing
It is as follows：
Wherein, u2Represent the image after noise reducing deblurring, ux, uyThe component in image x, y directions, λ are represented respectively2The tables of ＞ 0
Show parameter.
Preferably, the gray level image to the noise reducing deblurring, which carries out binary conversion treatment, includes using Otsu algorithm
Binary conversion treatment is carried out to the gray level image of the noise reducing deblurring.
In addition, the embodiment of the present invention additionally provides a kind of image deblurring system applied to car plate detection, the system
Including：
Image collection module：For obtaining coloured image；
Image conversion module：For described coloured image to be converted into gray level image；
Image noise reduction sound module：For described gray level image to be carried out into noise reduction sonication, the gray-scale map of noise reducing is obtained
Picture；
Image deblurring module：For the gray level image of described noise reduction to be carried out into deblurring processing, obtain noise reducing and go
Fuzzy gray image；
Image binaryzation module：For the gray image of described noise reducing deblurring to be carried out into binary conversion treatment, obtain
Binary image.
Preferably, the image deblurring system of the car plate detection, it is characterised in that described to turn described coloured image
Gray level image is changed to, including to the coloured image using the progress gradation conversion processing of brightness value conversion regime, the brightness value
Conversion formula is as follows：
I=0.3 × R+0.59 × G+0.11 × B
Wherein, I represents the brightness value of gray level image, and R represents red, and G represents green, and B represents blueness.
Preferably, the image deblurring system of the car plate detection, it is characterised in that described to enter described gray level image
Row noise reduction sonication includes：
Noise reduction sonication is carried out to the gray level image using weighting total variation algorithm, the formula of total variation algorithm is weighted such as
Under：
Or：
Wherein, i, j represent that 1,2,3..., y represents signal respectively；
Further, a signal x with random noise is givenn, one is obtained close to xnSignal ynWith smaller
Total variation, can be weighed, E (x, y) algorithmic formula is as follows with fidelity bound term E (x, y)：
Wherein, n represents 1,2,3..., and the fidelity bound term E (x, y) combines weighting total variation algorithm, solves minimum
Value, formula is as follows：
E (x, y)+λ V (y)
Wherein, λ represents weight.
Preferably, the image deblurring system of the car plate detection, it is characterised in that the gray scale by described noise reduction
Image, which carries out deblurring processing, to be included：
To the fuzzy core φ of a transfer functionbEstimated, obtain the fuzzy core φ of point transfer functionb；
According to described transfer function fuzzy core φb, deblurring processing is carried out to the gray level image of the noise reducing；
The fuzzy core φ of described pair of point transfer functionbCarry out algorithm for estimating as follows：
Wherein, λ1＞ 0 represents parameter, u1Represent the image after noise reducing, φbPoint transmission fuzzy core is represented, C1 represents area
Domain, L represents the length of side, I1Represent original image；
Further, on the basis of variational problem is strict convex surface, the fuzzy core φ of described transfer functionbIt is unique minimum
Value point is exactly the solution of Euler-Lagrange equation：
-Δφb+λ1(φ**I1-u1)*I1=0
Wherein, Δ φbRepresent that electrical transmission obscures the difference of sum；
According to described transfer function fuzzy core φb, deblurring processing, algorithm are carried out to the gray level image of the noise reducing
It is as follows：
Wherein, u2Represent the image after noise reducing deblurring, ux, uyThe component in image x, y directions, λ are represented respectively2The tables of ＞ 0
Show parameter.
Preferably, the image deblurring system of the car plate detection, it is characterised in that described by described noise reducing mould from
The gray image of paste, which carries out binary conversion treatment, to be included carrying out two to the gray level image of the noise reducing deblurring using Otsu algorithm
Value is handled.
Carried out successively by the image to car plate detection at gradation conversion, noise reduction sonication, deblurring processing and binaryzation
The binary image of the high accuracy obtained is managed, with more preferable effect.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the accompanying drawing used required in technology description to be briefly described, it is clear that, drawings in the following description are only this
Some embodiments of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can be with
Other accompanying drawings are obtained according to these accompanying drawings.
Fig. 1 is a kind of schematic flow sheet of image deblurring method applied to car plate detection of the embodiment of the present invention；
Fig. 2 is a kind of structural representation of image deblurring system applied to car plate detection of the embodiment of the present invention.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.It is based on
Embodiment in the present invention, it is all other that those of ordinary skill in the art are obtained under the premise of creative work is not made
Embodiment, belongs to the scope of protection of the invention.
Fig. 1 is a kind of schematic flow sheet of image deblurring method applied to car plate detection of the embodiment of the present invention, such as Fig. 1
Shown, this method includes：
S11：Obtain coloured image；
S12：Gradation conversion is carried out to coloured image, gray level image is obtained；
S13：Noise reduction sonication is carried out to gray level image, the gray level image of noise reducing is obtained；
S14：Deblurring processing is carried out to the gray level image of noise reducing, the gray level image of noise reducing deblurring is obtained；
S15：Binary conversion treatment is carried out to the gray level image of noise reducing deblurring, bianry image is obtained.
S11 is further illustrated：
By camera, the car plate of high-speed mobile automobile is shot, image is obtained from camera, the image is coloured image,
That is RGB image.
S12 is further illustrated：
Gray level image is image of the brightness value between 0 to 255, it is assumed that R, G, B are respectively the red, green, blue of coloured image
Component, the brightness value of gray level image is set to I, then the S11 coloured images obtained can be changed according to the following formula：
I=0.3 × R+0.59 × G+0.11 × B
Wherein, I represents the brightness value of gray level image, and R represents red, and G represents green, and B represents blueness, so as to get
State the gray level image of coloured image.
S13 is further illustrated：
Due to obtaining during image, may by electric current, unstable, camera is of low quality etc. that factor is influenceed, image can go out
Existing noise spot, uses weighting total variation method denoising, it is possible to reduce the interference of noise spot, and it is based on so to weight total variation method
Principle：With excessive and be probably that the signal of false detail has the integration of high total change, the i.e. absolute gradient of signal high.
Around this principle, total change of reduction signal eliminates unnecessary details with primary signal close to matching, while retaining all
Such as the material particular at edge.
Noise reduction sonication is carried out to S12 gray level image using weighting total variation algorithm, the formula of total variation algorithm is weighted
It is as follows：
Or：
Wherein, i, j represent that 1,2,3..., y represents signal respectively.
If giving a signal x with random noisen, one is obtained close to xnSignal ynWith smaller total
Variation, can be used to lower fidelity bound term E (x, y) to weigh,
Wherein, n represents 1,2,3....
Problem can just be changed into seeking following formula minimum value：
E (x, y)+λ V (y)
Wherein, λ represents weight, and this formula carries out seeking partial derivative to y, can construct Lagrange's equation to solve.
S14 is further illustrated：
In optical system, because the superposition of image is linear, therefore below equation is met：
Image (Object_x1, Object_y1)=Image (M*Object_x2, M*Object_y2)
Therefore, the image of denoising meets below equation：
I2=N (φb*I1)
Wherein, I2For the image of denoising, φbFor the fuzzy core of a transfer function, N is that noise produces function, I1For artwork
Picture.
To the fuzzy core φ of above-mentioned transfer functionbEstimated, obtain the fuzzy core φ of point transfer functionb：
Wherein, λ1＞ 0 represents parameter, u1Represent the image after noise reducing, φbPoint transmission fuzzy core is represented, C1 represents area
Domain, L represents the length of side, I1Represent original image；
Further, because variational problem is strict convex surface, then the fuzzy core φ of transfer function is putbUnique minimum point
It is exactly the solution of Euler-Lagrange equation：
-Δφb+λ1(φ**I1-u1)*I1=0
Wherein, Δ φbRepresent that electrical transmission obscures the difference of sum, λ1＞ 0 represents parameter, I1Represent original image, u1Represent noise reduction
Image after sound；
According to above-mentioned transfer function fuzzy core φb, deblurring processing is carried out to the gray level image of noise reducing, algorithm is such as
Under：
Wherein, u2Represent the image after noise reducing deblurring, ux, uyThe component in image x, y directions, λ are represented respectively2The tables of ＞ 0
Show parameter.
S15 is further illustrated：
Gray level image progress binary conversion treatment to noise reducing deblurring is handled using Otsu algorithm.
Fig. 2 is a kind of structural representation of image deblurring system applied to car plate detection of the embodiment of the present invention, such as Fig. 2
Shown, the system includes：
11：Image collection module, for obtaining coloured image；
12：Image conversion module, for described coloured image to be converted into gray level image；
13：Image noise reduction sound module, for described gray level image to be carried out into noise reduction sonication, obtains the gray scale of noise reducing
Image；
14：Image deblurring module, for the gray level image of described noise reduction to be carried out into deblurring processing, obtains noise reducing
The gray image of deblurring；
15：Image binaryzation module, for the gray image of described noise reducing deblurring to be carried out into binary conversion treatment, is obtained
Take binary image.
Further illustrated to 11：
Image collection module shoots the car plate of high-speed mobile automobile by camera, and the image got from camera is
Coloured image, i.e. RGB image.
Further illustrated to 12：
The coloured image that 11 image collection modules are obtained is converted to gray level image by image conversion module, and gray level image is bright
Image of the angle value between 0 to 255, it is assumed that R, G, B are respectively the red, green, blue component of coloured image, the brightness value of gray level image
I is set to, then the S11 coloured images obtained can be changed according to the following formula：
I=0.3 × R+0.59 × G+0.11 × B
Wherein, I represents the brightness value of gray level image, and R represents red, and G represents green, and B represents blueness, so as to get
State the gray level image of coloured image.
Further illustrated to 13：
The gray level image of 12 image conversion modules is carried out noise reduction sonication by image noise reduction sound module, because 11 images are obtained
When module obtains image, may by electric current, unstable, camera is of low quality etc. that factor is influenceed, noise occurs in image
Point, uses weighting total variation method denoising, it is possible to reduce the interference of noise spot, and it is based on such former to weight total variation method
Reason：With excessive and be probably that the signal of false detail has the integration of high total change, the i.e. absolute gradient of signal high.According to
This principle, total change of reduction signal, close to matching, eliminates unnecessary details, while retaining such as side with primary signal
The material particular of edge.
Noise reduction sonication is carried out to 12 gray level image using weighting total variation algorithm, the formula of total variation algorithm is weighted such as
Under：
Or：
Wherein, i, j represent that 1,2,3..., y represents signal respectively.
If giving a signal x with random noisen, one is obtained close to xnSignal ynWith smaller total
Variation, can be used to lower fidelity bound term E (x, y) to weigh,
Wherein, n represents 1,2,3....
Problem can just be changed into seeking following formula minimum value：
E (x, y)+λ V (y)
Wherein, λ represents weight, and this formula carries out seeking partial derivative to y, can construct Lagrange's equation to solve.
Further illustrated to 14：
Gray level image after the noise reducing of 13 image noise reduction sound modules is carried out deblurring processing by image deblurring module,
In optical system, because the superposition of image is linear, therefore below equation is met：
Image (Object_x1, Object_y1)=Image (M*Object_x2, M*Object_y2)
Therefore, the image of denoising meets below equation：
I2=N (φb*I1)
Wherein, I2For the image of denoising, φbFor the fuzzy core of a transfer function, N is that noise produces function, I1For artwork
Picture.
To the fuzzy core φ of above-mentioned transfer functionbEstimated, obtain the fuzzy core φ of point transfer functionb：
Wherein, λ1＞ 0 represents parameter, u1Represent the image after noise reducing, φbPoint transmission fuzzy core is represented, C1 represents area
Domain, L represents the length of side, I1Represent original image；
Further, because variational problem is strict convex surface, then the fuzzy core φ of transfer function is putbUnique minimum point
It is exactly the solution of Euler-Lagrange equation：
-Δφb+λ1(φ**I1-u1)*I1=0
Wherein, Δ φbRepresent that electrical transmission obscures the difference of sum, λ1＞ 0 represents parameter, I1Represent original image, u1Represent noise reduction
Image after sound；
According to above-mentioned transfer function fuzzy core φb, deblurring processing is carried out to the gray level image of noise reducing, algorithm is such as
Under：
Wherein, u2Represent the image after noise reducing deblurring, ux, uyThe component in image x, y directions, λ are represented respectively2The tables of ＞ 0
Show parameter.
Further illustrated to 15：
Image binaryzation module carries out two-value to the gray level image after the noise reducing deblurring of 14 noise reducing deblurring modules
Change processing to be handled using Otsu algorithm.
One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment is can
To instruct the hardware of correlation to complete by program, the program can be stored in a computer-readable recording medium, storage
Medium can include：Read-only storage (ROM, ReadOnly Memory), random access memory (RAM, Random Access
Memory), disk or CD etc..
In addition, a kind of image deblurring method applied to car plate detection for being provided above the embodiment of the present invention and being
System is described in detail, and specific case should be employed herein the principle and embodiment of the present invention are set forth, with
The explanation of upper embodiment is only intended to the method and its core concept for helping to understand the present invention；Simultaneously for the general of this area
Technical staff, according to the thought of the present invention, will change in specific embodiments and applications, in summary,
This specification content should not be construed as limiting the invention.
Claims (10)
1. a kind of image deblurring method applied to car plate detection, it is characterised in that methods described includes：
Obtain coloured image；
Gradation conversion is carried out to the coloured image, gray level image is obtained；
Noise reduction sonication is carried out to the gray level image, the gray level image of noise reducing is obtained；
Deblurring processing is carried out to the gray level image of the noise reducing, the gray level image of noise reducing deblurring is obtained；
Binary conversion treatment is carried out to the gray level image of the noise reducing deblurring, bianry image is obtained.
2. the image deblurring method of car plate detection according to claim 1, it is characterised in that described to the coloured image
Gradation conversion is carried out, including to the coloured image using the progress gradation conversion processing of brightness value conversion regime, the brightness value
Conversion formula is as follows：
I=0.3 × R+0.59 × G+0.11 × B
Wherein, I represents the brightness value of gray level image, and R represents red, and G represents green, and B represents blueness.
3. the image deblurring method of car plate detection according to claim 1, it is characterised in that described to the gray level image
Carrying out noise reduction sonication includes：
Noise reduction sonication is carried out to the gray level image using weighting total variation algorithm, the formula for weighting total variation algorithm is as follows：
<mrow>
<mi>V</mi>
<mrow>
<mo>(</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</munder>
<msqrt>
<mrow>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
</mrow>
Or：
<mrow>
<mi>V</mi>
<mrow>
<mo>(</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</munder>
<msqrt>
<mrow>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</munder>
<mo>|</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>|</mo>
<mo>+</mo>
<mo>|</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>|</mo>
</mrow>
Wherein, i, j represent that 1,2,3..., y represents signal respectively；
Further, a signal x with random noise is givenn, one is obtained close to xnSignal ynWith smaller total
Variation, can be weighed, E (x, y) algorithmic formula is as follows with fidelity bound term E (x, y)：
<mrow>
<mi>E</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mn>2</mn>
</mfrac>
<munder>
<mo>&Sigma;</mo>
<mi>n</mi>
</munder>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>n</mi>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mi>n</mi>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
Wherein, n represents 1,2,3..., and the fidelity bound term E (x, y) combines weighting total variation algorithm, solves minimum value, public
Formula is as follows：
E (x, y)+λ V (y)
Wherein, λ represents weight.
4. the image deblurring method of car plate detection according to claim 1, it is characterised in that described to the noise reducing
Gray level image carries out deblurring processing, including：
To the fuzzy core φ of a transfer functionbEstimated, obtain the fuzzy core φ of point transfer functionb；
According to described transfer function fuzzy core φb, deblurring processing is carried out to the gray level image of the noise reducing；
The fuzzy core φ of described pair of point transfer functionbCarry out algorithm for estimating as follows：
<mrow>
<msub>
<mi>&phi;</mi>
<mi>b</mi>
</msub>
<mo>=</mo>
<msub>
<mi>argmin</mi>
<mi>&phi;</mi>
</msub>
<mfrac>
<mn>1</mn>
<mn>2</mn>
</mfrac>
<msub>
<mo>&Integral;</mo>
<mrow>
<mi>C</mi>
<mn>1</mn>
</mrow>
</msub>
<mo>|</mo>
<mo>&dtri;</mo>
<mi>&phi;</mi>
<msup>
<mo>|</mo>
<mn>2</mn>
</msup>
<mo>+</mo>
<mfrac>
<msub>
<mi>&lambda;</mi>
<mn>1</mn>
</msub>
<mn>2</mn>
</mfrac>
<mo>|</mo>
<mo>|</mo>
<mi>&phi;</mi>
<mo>*</mo>
<msub>
<mi>I</mi>
<mn>1</mn>
</msub>
<mo>-</mo>
<msub>
<mi>u</mi>
<mn>1</mn>
</msub>
<mo>|</mo>
<msubsup>
<mo>|</mo>
<mrow>
<msup>
<mi>L</mi>
<mn>2</mn>
</msup>
<mrow>
<mo>(</mo>
<mi>C</mi>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
<mn>2</mn>
</msubsup>
</mrow>
Wherein, λ1＞ 0 represents parameter, u1Represent the image after noise reducing, φbPoint transmission fuzzy core is represented, C1 represents region, L tables
Show the length of side, I1Represent original image；
Further, on the basis of variational problem is strict convex surface, the fuzzy core φ of described transfer functionbUnique minimum point
It is exactly the solution of Euler-Lagrange equation：
-Δφb+λ1(φ**I1-u1)*I1=0
Wherein, Δ φbRepresent that electrical transmission obscures the difference of sum；
According to described transfer function fuzzy core φb, deblurring processing is carried out to the gray level image of the noise reducing, algorithm is as follows：
<mrow>
<msub>
<mi>u</mi>
<mn>2</mn>
</msub>
<mo>=</mo>
<msub>
<mi>argmin</mi>
<mi>u</mi>
</msub>
<mo>&Integral;</mo>
<mo>|</mo>
<msub>
<mi>u</mi>
<mi>x</mi>
</msub>
<mo>|</mo>
<mo>+</mo>
<mo>|</mo>
<msub>
<mi>u</mi>
<mi>y</mi>
</msub>
<mo>|</mo>
<mo>+</mo>
<mfrac>
<msub>
<mi>&lambda;</mi>
<mn>2</mn>
</msub>
<mn>2</mn>
</mfrac>
<mo>|</mo>
<mo>|</mo>
<msub>
<mi>&phi;</mi>
<mi>b</mi>
</msub>
<mo>*</mo>
<msub>
<mi>I</mi>
<mn>1</mn>
</msub>
<mo>-</mo>
<msub>
<mi>u</mi>
<mn>1</mn>
</msub>
<mo>|</mo>
<msubsup>
<mo>|</mo>
<msup>
<mi>L</mi>
<mn>2</mn>
</msup>
<mn>2</mn>
</msubsup>
</mrow>
Wherein, u2Represent the image after noise reducing deblurring, ux, uyThe component in image x, y directions, λ are represented respectively2＞ 0 represents ginseng
Number.
5. the image deblurring method of car plate detection according to claim 1, it is characterised in that described to be gone to the noise reducing
Fuzzy gray level image, which carries out binary conversion treatment, to be included carrying out the gray level image of the noise reducing deblurring using Otsu algorithm
Binary conversion treatment.
6. a kind of image deblurring system applied to car plate detection, it is characterised in that the system includes：
Image collection module：For obtaining coloured image；
Image conversion module：For described coloured image to be converted into gray level image；
Image noise reduction sound module：For described gray level image to be carried out into noise reduction sonication, the gray level image of noise reducing is obtained；
Image deblurring module：For the gray level image of described noise reduction to be carried out into deblurring processing, noise reducing deblurring is obtained
Gray image；
Image binaryzation module：For the gray image of described noise reducing deblurring to be carried out into binary conversion treatment, two-value is obtained
Change image.
7. the image deblurring system of car plate detection according to claim 6, it is characterised in that described by described cromogram
It is described bright as being converted to gray level image, including to the coloured image using the progress gradation conversion processing of brightness value conversion regime
Angle value conversion formula is as follows：
I=0.3 × R+0.59 × G+0.11 × B
Wherein, I represents the brightness value of gray level image, and R represents red, and G represents green, and B represents blueness.
8. the image deblurring system of car plate detection according to claim 6, it is characterised in that described by described gray-scale map
Include as carrying out noise reduction sonication：
Noise reduction sonication is carried out to the gray level image using weighting total variation algorithm, the formula for weighting total variation algorithm is as follows：
<mrow>
<mi>V</mi>
<mrow>
<mo>(</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</munder>
<msqrt>
<mrow>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
</mrow>
Or：
<mrow>
<mi>V</mi>
<mrow>
<mo>(</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</munder>
<msqrt>
<mrow>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</munder>
<mo>|</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>|</mo>
<mo>+</mo>
<mo>|</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>|</mo>
</mrow>
Wherein, i, j represent that 1,2,3..., y represents signal respectively；
Further, a signal x with random noise is givenn, one is obtained close to xnSignal ynWith smaller total
Variation, can be weighed, E (x, y) algorithmic formula is as follows with fidelity bound term E (x, y)：
<mrow>
<mi>E</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mn>2</mn>
</mfrac>
<munder>
<mo>&Sigma;</mo>
<mi>n</mi>
</munder>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>n</mi>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mi>n</mi>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
Wherein, n represents 1,2,3..., and the fidelity bound term E (x, y) combines weighting total variation algorithm, solves minimum value, public
Formula is as follows：
E (x, y)+λ V (y)
Wherein, λ is weight.
9. the image deblurring system of car plate detection according to claim 6, it is characterised in that described by described noise reduction
Gray level image, which carries out deblurring processing, to be included：
To the fuzzy core φ of a transfer functionbEstimated, obtain the fuzzy core φ of point transfer functionb；
According to described transfer function fuzzy core φb, deblurring processing is carried out to the gray level image of the noise reducing；
The fuzzy core φ of described pair of point transfer functionbCarry out algorithm for estimating as follows：
<mrow>
<msub>
<mi>&phi;</mi>
<mi>b</mi>
</msub>
<mo>=</mo>
<msub>
<mi>argmin</mi>
<mi>&phi;</mi>
</msub>
<mfrac>
<mn>1</mn>
<mn>2</mn>
</mfrac>
<msub>
<mo>&Integral;</mo>
<mrow>
<mi>C</mi>
<mn>1</mn>
</mrow>
</msub>
<mo>|</mo>
<mo>&dtri;</mo>
<mi>&phi;</mi>
<msup>
<mo>|</mo>
<mn>2</mn>
</msup>
<mo>+</mo>
<mfrac>
<msub>
<mi>&lambda;</mi>
<mn>1</mn>
</msub>
<mn>2</mn>
</mfrac>
<mo>|</mo>
<mo>|</mo>
<mi>&phi;</mi>
<mo>*</mo>
<msub>
<mi>I</mi>
<mn>1</mn>
</msub>
<mo>-</mo>
<msub>
<mi>u</mi>
<mn>1</mn>
</msub>
<mo>|</mo>
<msubsup>
<mo>|</mo>
<mrow>
<msup>
<mi>L</mi>
<mn>2</mn>
</msup>
<mrow>
<mo>(</mo>
<mi>C</mi>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
<mn>2</mn>
</msubsup>
</mrow>
Wherein, λ1＞ 0 represents parameter, u1Represent the image after noise reducing, φbPoint transmission fuzzy core is represented, C1 represents region, L tables
Show the length of side, I1Represent original image；
Further, on the basis of variational problem is strict convex surface, the fuzzy core of described transfer functionUnique minimum point
It is exactly the solution of Euler-Lagrange equation：
-Δφb+λ1(φ**I1-u1)*I1=0
Wherein, Δ φbRepresent that electrical transmission obscures the difference of sum；
According to described transfer function fuzzy coreDeblurring processing is carried out to the gray level image of the noise reducing, algorithm is as follows：
<mrow>
<msub>
<mi>u</mi>
<mn>2</mn>
</msub>
<mo>=</mo>
<msub>
<mi>argmin</mi>
<mi>u</mi>
</msub>
<mo>&Integral;</mo>
<mo>|</mo>
<msub>
<mi>u</mi>
<mi>x</mi>
</msub>
<mo>|</mo>
<mo>+</mo>
<mo>|</mo>
<msub>
<mi>u</mi>
<mi>y</mi>
</msub>
<mo>|</mo>
<mo>+</mo>
<mfrac>
<msub>
<mi>&lambda;</mi>
<mn>2</mn>
</msub>
<mn>2</mn>
</mfrac>
<mo>|</mo>
<mo>|</mo>
<msub>
<mi>&phi;</mi>
<mi>b</mi>
</msub>
<mo>*</mo>
<msub>
<mi>I</mi>
<mn>1</mn>
</msub>
<mo>-</mo>
<msub>
<mi>u</mi>
<mn>1</mn>
</msub>
<mo>|</mo>
<msubsup>
<mo>|</mo>
<msup>
<mi>L</mi>
<mn>2</mn>
</msup>
<mn>2</mn>
</msubsup>
</mrow>
Wherein, u2Represent the image after noise reducing deblurring, ux, uyThe component in image x, y directions, λ are represented respectively2＞ 0 represents ginseng
Number.
10. the image deblurring system of car plate detection according to claim 6, it is characterised in that described by described noise reduction
The gray image of sound deblurring, which carries out binary conversion treatment, includes the gray level image using Otsu algorithm to the noise reducing deblurring
Carry out binary conversion treatment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710316939.2A CN107194887A (en) | 2017-05-08 | 2017-05-08 | A kind of image deblurring method and system applied to car plate detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710316939.2A CN107194887A (en) | 2017-05-08 | 2017-05-08 | A kind of image deblurring method and system applied to car plate detection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107194887A true CN107194887A (en) | 2017-09-22 |
Family
ID=59873751
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710316939.2A Pending CN107194887A (en) | 2017-05-08 | 2017-05-08 | A kind of image deblurring method and system applied to car plate detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107194887A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050047672A1 (en) * | 2003-06-17 | 2005-03-03 | Moshe Ben-Ezra | Method for de-blurring images of moving objects |
| CN103455994A (en) * | 2012-05-28 | 2013-12-18 | 佳能株式会社 | Method and equipment for determining image blurriness |
| CN104331871A (en) * | 2014-12-02 | 2015-02-04 | 苏州大学 | Image de-blurring method and image de-blurring device |
| CN106339996A (en) * | 2016-09-09 | 2017-01-18 | 江南大学 | Image blind defuzzification method based on hyper-Laplacian prior |
-
2017
- 2017-05-08 CN CN201710316939.2A patent/CN107194887A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050047672A1 (en) * | 2003-06-17 | 2005-03-03 | Moshe Ben-Ezra | Method for de-blurring images of moving objects |
| CN103455994A (en) * | 2012-05-28 | 2013-12-18 | 佳能株式会社 | Method and equipment for determining image blurriness |
| CN104331871A (en) * | 2014-12-02 | 2015-02-04 | 苏州大学 | Image de-blurring method and image de-blurring device |
| CN106339996A (en) * | 2016-09-09 | 2017-01-18 | 江南大学 | Image blind defuzzification method based on hyper-Laplacian prior |
Non-Patent Citations (1)
| Title |
|---|
| 王超 等: "基于L0 正则化的车牌图像去模糊", 《电子设计工程》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106056155B (en) | Superpixel segmentation method based on boundary information fusion | |
| CN103605977B (en) | Extracting method of lane line and device thereof | |
| CN110796595B (en) | Tone mapping method and device and electronic equipment | |
| CN110163801B (en) | Image super-resolution and coloring method, system and electronic equipment | |
| CN105374013B (en) | Image visibility restorative procedure and image processing apparatus based on double dark channel priors | |
| Luan et al. | Fast single image dehazing based on a regression model | |
| CN107633220A (en) | A kind of vehicle front target identification method based on convolutional neural networks | |
| CN106570474A (en) | Micro expression recognition method based on 3D convolution neural network | |
| CN101425179B (en) | Face image relighting method and device | |
| WO2024051296A1 (en) | Method and apparatus for obstacle detection in complex weather | |
| CN104182947A (en) | Low-illumination image enhancement method and system | |
| CN106664368A (en) | Image processing apparatus, image processing method, recording medium, and program | |
| CN112070158A (en) | Facial flaw detection method based on convolutional neural network and bilateral filtering | |
| CN111209858A (en) | Real-time license plate detection method based on deep convolutional neural network | |
| CN104021531A (en) | Improved method for enhancing dark environment images on basis of single-scale Retinex | |
| CN103366390A (en) | Terminal, image processing method and device thereof | |
| CN113066025A (en) | Image defogging method based on incremental learning and feature and attention transfer | |
| CN103489168A (en) | Enhancing method and system for infrared image being converted to pseudo color image in self-adaptive mode | |
| CN102999890A (en) | Method for dynamically calibrating image light intensity on basis of environmental factors | |
| CN104616259B (en) | A kind of adaptive non-local mean image de-noising method of noise intensity | |
| CN114862707A (en) | Multi-scale feature recovery image enhancement method and device and storage medium | |
| CN112991236B (en) | Image enhancement method and device based on template | |
| CN107451986A (en) | Single infrared image enhancement method based on fusion technology | |
| CN112288622B (en) | Multi-scale generation countermeasure network-based camouflaged image generation method | |
| CN107194887A (en) | A kind of image deblurring method and system applied to car plate detection |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170922 |
|
| WD01 | Invention patent application deemed withdrawn after publication |