CN107451978A - A kind of image processing method, device and equipment - Google Patents
A kind of image processing method, device and equipment Download PDFInfo
- Publication number
- CN107451978A CN107451978A CN201710666951.6A CN201710666951A CN107451978A CN 107451978 A CN107451978 A CN 107451978A CN 201710666951 A CN201710666951 A CN 201710666951A CN 107451978 A CN107451978 A CN 107451978A
- Authority
- CN
- China
- Prior art keywords
- pixel
- image
- denoising
- parameter
- invalid
- 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.)
- Granted
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 23
- 238000000034 method Methods 0.000 claims abstract description 20
- 238000012545 processing Methods 0.000 claims description 19
- 230000009466 transformation Effects 0.000 claims description 12
- 238000009826 distribution Methods 0.000 claims description 8
- 241000208340 Araliaceae Species 0.000 claims description 4
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims description 4
- 235000003140 Panax quinquefolius Nutrition 0.000 claims description 4
- 235000008434 ginseng Nutrition 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 description 16
- 230000000694 effects Effects 0.000 description 12
- 238000013461 design Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000005457 optimization Methods 0.000 description 4
- 230000000717 retained effect Effects 0.000 description 4
- 238000003860 storage Methods 0.000 description 4
- 238000003384 imaging method Methods 0.000 description 3
- 238000002595 magnetic resonance imaging Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000012827 research and development Methods 0.000 description 3
- 238000004590 computer program Methods 0.000 description 2
- 239000004744 fabric Substances 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000003213 activating effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
Classifications
-
- G06T5/70—
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20048—Transform domain processing
- G06T2207/20056—Discrete and fast Fourier transform, [DFT, FFT]
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20224—Image subtraction
-
- 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
- G06T2207/30004—Biomedical image processing
Abstract
The embodiment of the present application discloses a kind of image processing method, device and equipment, wherein, method includes：Obtain differential image, differential image is the image that original image and Pre-denoised image subtract each other to obtain, Pre-denoised image is image of the original image after predenoising, the pixel of each differential image is corresponding with weight respectively, and the pixel of differential image includes effective denoising pixel and invalid denoising pixel；The random value for determining denoising parameter corresponding to differential image pixel, denoising parameter include the first parameter, and/or the second parameter, and/or the 3rd parameter；First parameter is the weight corresponding to effective denoising pixel in default proportion range, ratio of second parameter between the number of effective denoising pixel and the number of invalid denoising pixel, the 3rd parameter are the position of effective denoising pixel and/or invalid denoising pixel；According to the pixel value of differential image pixel and the value of the denoising parameter determined at random, denoising is carried out to original image or Pre-denoised image.
Description
Technical field
The application is related to image processing field, more particularly to a kind of image processing method, device and equipment.
Background technology
Noise refers to the error in image, and in medical domain, picture noise can be that white noise, motion artifacts, reconstruction are drawn
Error entered etc..Denoising is to acquire the post processing link for being generally required for carrying out after image.The purpose of denoising is balances noise
Removal and image information protection.Because if noise remove is excessive, then a part of real image information may be lost；
And noise remove is very few, then the effect of denoising is not reached.The removal of noise for each pixel in image generally by distinguishing
Corresponding weight is multiplied by realize, therefore the design of weight is very crucial for the removal effect of noise.
However, because noise is uneven on each pixel, so even if giving the overall letter of image after processing
Make an uproar and compare, can not also be accurately obtained weight corresponding to each pixel.
The content of the invention
In order to solve technical problem present in prior art, this application provides a kind of image processing method, device and
Equipment, realize that balances noise removes and retained the purpose of image information.
A kind of image processing method is present embodiments provided, methods described includes：
Differential image is obtained, the differential image is the image that original image and Pre-denoised image subtract each other to obtain, described pre- to go
Image of making an uproar is image of the original image after predenoising, and the pixel of each differential image corresponds to respectively has the right
Weight, the pixel of the differential image include effective denoising pixel and invalid denoising pixel, the effectively denoising pixel
It is not the pixel of default weight for corresponding weight, the invalid denoising pixel is the picture that corresponding weight is default weight
Vegetarian refreshments；
The value of denoising parameter corresponding to the differential image pixel is determined at random, and the denoising parameter includes the first ginseng
Number, and/or the second parameter, and/or the 3rd parameter；Wherein, first parameter is effectively gone to be described in default proportion range
Make an uproar weight corresponding to pixel, second parameter is the effectively number of denoising pixel and of invalid denoising pixel
Ratio between number, the 3rd parameter are the effectively position of denoising pixel and/or the invalid denoising pixel；
According to the pixel value of the differential image pixel and the value of the denoising parameter determined at random, to the original
Image or the Pre-denoised image carry out denoising, obtain target image.
Optionally, this method also includes：
Obtain the noise information of the original image；
The differential image is optimized according to the noise information of the original image.
Optionally, the noise information includes following at least one：
Area-of-interest and noise intensity probability distribution.
Optionally, the acquisition differential image includes：
Time domain original image and time domain Pre-denoised image are obtained, the time domain original image is the original that picture signal is time-domain signal
Image, time domain Pre-denoised image are the Pre-denoised images that picture signal is time-domain signal；
The time domain original image is subjected to Fourier transformation, obtains frequency domain original image, the frequency domain original image is image letter
Number be frequency-region signal original image；
The time domain Pre-denoised image is subjected to Fourier transformation, obtains frequency domain Pre-denoised image, the frequency domain predenoising
Image is the Pre-denoised image that picture signal is frequency-region signal；
The frequency domain original image and the frequency domain Pre-denoised image are subtracted each other, obtain the differential image；
It is described that the original image or Pre-denoised image progress denoising are included：
Denoising is carried out to the frequency domain original image or the frequency domain Pre-denoised image.
Optionally, the value for determining denoising parameter corresponding to the differential image pixel at random includes：
The effectively ratio of the number of denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
The position of point and/or invalid denoising pixel is fixed, and weight corresponding to effective denoising pixel is random in default proportion range；
Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
And/or the position of invalid denoising pixel is random, weight corresponding to effective denoising pixel is random in default proportion range；Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is random, effective denoising pixel
And/or the position of invalid denoising pixel is random, weight corresponding to effective denoising pixel is fixed；Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
And/or the position of invalid denoising pixel is random, weight corresponding to effective denoising pixel is fixed.
The embodiment of the present application provides a kind of image processing apparatus, and described device includes：
Differential image acquiring unit, determining unit and denoising unit；
Wherein, the differential image acquiring unit, for obtaining differential image, the differential image is original image and gone in advance
The image that image subtraction of making an uproar obtains, the Pre-denoised image are image of the original image after predenoising, each difference
The pixel of different image is all corresponding with weight respectively, and the pixel of the differential image includes effective denoising pixel and invalid gone
Make an uproar pixel, the effectively denoising pixel be corresponding to weight be not default weight pixel, the invalid denoising pixel
Point is to preset the pixel of weight for corresponding weight；
The determining unit, it is described to go for determining the value of denoising parameter corresponding to the differential image pixel at random
Parameter of making an uproar includes the first parameter, and/or the second parameter, and/or the 3rd parameter；Wherein, first parameter is in default weight
In the range of the weight effectively corresponding to denoising pixel, second parameter for the effectively denoising pixel number and nothing
Imitate the ratio between the number of denoising pixel, the 3rd parameter is the effectively denoising pixel and/or described invalid goes
Make an uproar the position of pixel；
The denoising unit, for the pixel value according to the differential image pixel and the denoising determined at random
The value of parameter, denoising is carried out to the original image or the Pre-denoised image, obtains target image.
Optionally, described device also includes：
Noise information acquiring unit and denoising subelement；
The noise information acquiring unit, for obtaining the noise information of the original image；
The denoising subelement, for optimizing the differential image according to the noise information of the original image.
Optionally, the differential image acquiring unit includes：
It is single that time-domain image acquiring unit, original image converter unit, Pre-denoised image converter unit and differential image obtain son
Member；
Wherein, the time-domain image acquiring unit, for obtaining time domain original image and time domain Pre-denoised image, the time domain
Original image is the original image that picture signal is time-domain signal, and time domain Pre-denoised image is the predenoising that picture signal is time-domain signal
Image；
The original image converter unit, for the time domain original image to be carried out into Fourier transformation, frequency domain original image is obtained,
The frequency domain original image is the original image that picture signal is frequency-region signal；
The Pre-denoised image converter unit, for the time domain Pre-denoised image to be carried out into Fourier transformation, obtain frequency
Domain Pre-denoised image, the frequency domain Pre-denoised image are the Pre-denoised images that picture signal is frequency-region signal；
The differential image obtains subelement, for the frequency domain original image and the frequency domain Pre-denoised image to be subtracted each other,
Obtain the differential image；
The denoising unit, is specifically used for：
According to the pixel value and weight of the differential image pixel, the frequency domain original image or the frequency domain are gone in advance
Image of making an uproar carries out denoising, obtains target image.
Optionally, the determining unit, is specifically used for：
The effectively ratio of the number of denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
The position of point and/or invalid denoising pixel is fixed, and weight corresponding to effective denoising pixel is random in default proportion range；
Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
And/or the position of invalid denoising pixel is random, weight corresponding to effective denoising pixel is random in default proportion range；Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is random, effective denoising pixel
And/or the position of invalid denoising pixel is random, weight corresponding to effective denoising pixel is fixed；Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
And/or the position of invalid denoising pixel is random, weight corresponding to effective denoising pixel is fixed.
The embodiment of the present application provides a kind of image processing equipment, and the equipment includes：Processor, for storing the place
Manage the memory and display of device executable instruction；
Wherein, the processor is configured as：
Differential image is obtained, the differential image is the image that original image and Pre-denoised image subtract each other to obtain, described pre- to go
Image of making an uproar is image of the original image after predenoising, and the pixel of each differential image corresponds to respectively has the right
Weight, the pixel of the differential image include effective denoising pixel and invalid denoising pixel, the effectively denoising pixel
It is not the pixel of default weight for corresponding weight, the invalid denoising pixel is the picture that corresponding weight is default weight
Vegetarian refreshments；
The value of denoising parameter corresponding to the differential image pixel is determined at random, and the denoising parameter includes the first ginseng
Number, and/or the second parameter, and/or the 3rd parameter；Wherein, first parameter is effectively gone to be described in default proportion range
Make an uproar weight corresponding to pixel, second parameter is the effectively number of denoising pixel and of invalid denoising pixel
Ratio between number, the 3rd parameter are the effectively position of denoising pixel and/or the invalid denoising pixel；
According to the pixel value of the differential image pixel and the value of the denoising parameter determined at random, to the original
Image or the Pre-denoised image carry out denoising, obtain target image；
The display, for showing the target image.
The application determines denoising parameter corresponding to the differential image pixel at random by obtaining differential image
Value, the denoising parameter includes the first parameter, and/or the second parameter, and/or the 3rd parameter, finally according to the differential image
The value of the pixel value of pixel and the denoising parameter determined at random, the original image or the Pre-denoised image are carried out
Denoising, obtain target image., can be with if the target image generated at random is not the image that user intentionally gets
The target image of multiple different noise levels is generated by adjusting " slide bar ", therefrom selects the target image of a best results,
Realize that balances noise removes and retained the purpose of image information.
Brief description of the drawings
, below will be to embodiment or existing in order to illustrate more clearly of the embodiment of the present application or technical scheme of the prior art
There is the required accompanying drawing used in technology description to be briefly described, it should be apparent that, drawings in the following description are only this
Some embodiments described in application, for those of ordinary skill in the art, on the premise of not paying creative work,
Other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 is a kind of flow chart for image processing method that the embodiment of the present application provides；
Fig. 2 is that obtained target image schematic diagram is handled the white noise in image in the embodiment of the present application；
Fig. 3 is that in the embodiment of the present application the motion artifacts in image are handled with obtained target image schematic diagram；
Fig. 4 is the flow chart for another image processing method that the embodiment of the present application provides；
Fig. 5 is the flow chart for another image processing method that the embodiment of the present application provides；
Fig. 6 is a kind of structured flowchart for image processing apparatus that the embodiment of the present application provides；
Fig. 7 is a kind of hardware architecture diagram for image processing apparatus that the embodiment of the present application provides.
Embodiment
In computer picture field, image is embodied in a manner of signal, except carrying image sheet in picture signal
The information of body, such as brightness etc., also tend to carry noise.The size of noise influences the quality of image, and noise is excessive, and image is certainly
The information of body is just more difficult to be embodied；Noise is too small, and the presentation effect of the information of low contrast is poor in image, so image is gone
The process made an uproar is exactly the process of balances noise and image information.
Image is made up of each pixel, and the general principle of image denoising is exactly each pixel to be multiplied by respectively pair
The weight answered, how weight being designed, can reach and both protect image information, and can effectively removes unnecessary noise,
It is a problem of industry.Because for piece image, noise corresponding to each pixel is uneven, even if knowing
The target signal to noise ratio of whole image denoising, can not also know how many weight distributed respectively for each pixel on earth, can be only achieved
The target signal to noise ratio.
In order to solve above-mentioned problem, this application provides a kind of image processing method, device and equipment, its main thought is
Be each pixel Random Design weight on the premise of Constrained item, i.e., it is poor according to the value of default denoising parameter, random determination
Weight corresponding to the pixel of different image.After multiple processing are obtained according to the weight determined at random after image, can therefrom it select
Image is shown after selecting a best processing of effect.The above method is broken the normal procedure thinking, is no longer entangled with being each on earth
Pixel design how many weight, how the information of balances noise and image in itself, but pass through random manner generate target figure
Picture, therefrom the good target image of Selection effect come reach provide the preferable image of balanced degree purpose.
In order that those skilled in the art more fully understand application scheme, below in conjunction with the embodiment of the present application
Accompanying drawing, the technical scheme in the embodiment of the present application is clearly and completely described, it is clear that described embodiment is only this
Apply for part of the embodiment, rather than whole embodiments.Based on the embodiment in the application, those of ordinary skill in the art exist
The every other embodiment obtained under the premise of creative work is not made, belongs to the scope of the application protection.
Embodiment of the method：
Referring to Fig. 1, the figure is a kind of flow chart for image processing method that the embodiment of the present application provides.
The image processing method that the present embodiment provides comprises the following steps：
S101：Obtain differential image.
The purpose of the present embodiment is to obtain for original image, the target image of noise and image information balance.
In order to obtain the target image, the present embodiment also needs to obtain Pre-denoised image, and Pre-denoised image is that original image passes through predenoising
Image afterwards.There is certain difference in the image effect of original image and the image effect of Pre-denoised image, if original image is made an uproar
Sound is higher, then the noise of Pre-denoised image should be lower than the noise of original image；If the noise of original image is relatively low, image fault
It is more serious, then the noise of Pre-denoised image should be than artwork image height, so that the information of image is recovered.
In the present embodiment, original image and Pre-denoised image subtract each other, and obtain differential image.Original image and Pre-denoised image phase
The implication subtracted refers to the pixel value of original image pixel, and the pixel value of pixel subtracts each other in corresponding Pre-denoised image, obtains
To pixel value be differential image corresponding pixel points pixel value.
After differential image is obtained, target image can be obtained as follows：The pixel of each differential image
Weight is corresponding with respectively, and the pixel value of each pixel of differential image is multiplied by each self-corresponding weight respectively, and it is poor " after adjustment " to obtain
Pixel value corresponding to different each pixel of image, artwork image subtraction should obtain target image by differential image " after adjustment ", or go in advance
Image of making an uproar obtain target image plus differential image " after adjustment " is somebody's turn to do.As can be seen that how to obtain each pixel pair of differential image
The weight answered is key point.
The pixel of differential image can include effective denoising pixel and invalid denoising pixel, effective denoising pixel
It is not the pixel of default weight for corresponding weight, invalid denoising pixel is the pixel that corresponding weight is default weight
Point.The default weight can be, for example, 0 or 1.
S102：The value of denoising parameter corresponding to the differential image pixel is determined at random.
In the present embodiment, the denoising parameter includes the first parameter, and/or the second parameter, and/or the 3rd parameter.
Wherein, first parameter is the weight effectively corresponding to denoising pixel.The effectively denoising pixel pair
The weight answered can be with completely random, can also be random in default proportion range.Default proportion range defines effective denoising picture
The random scope of vegetarian refreshments, in order to which effective denoising pixel and invalid denoising pixel are made a distinction, the default proportion range
Can be not include the default weight.
Optionally, in order to ensure the picture quality of target image between original image and Pre-denoised image, if default power
Weight is 0, then the default proportion range be chosen as (0,1]；If default weight is 1, then the default proportion range be chosen as [0,
1)。
Second parameter is the effectively ratio of the number of denoising pixel and the number of invalid denoising pixel.It is false
If piece image has 300 pixels, effective denoising pixel has 250, and invalid denoising pixel has 50, then effectively goes
The ratio of pixel and invalid denoising pixel of making an uproar is 5:1.
3rd parameter is the position of effective denoising pixel and/or invalid denoising pixel.Pixel in differential image
Can be partly or entirely effective denoising pixel, but can not all invalid denoising pixels.
Several random cases that this step covers are described in detail below：
1st, the ratio of the number of effective denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
And/or the position of invalid denoising pixel is fixed, weight corresponding to effective denoising pixel is random in default proportion range.
In the present embodiment, fixed implication refers to preset, rather than random determination.
In this case, if effectively the position of denoising pixel or invalid denoising pixel determines, then effective denoising picture
Ratio between the number of vegetarian refreshments and the number of invalid denoising pixel has been inherently derived determination, and random object is effective denoising
Weight corresponding to pixel.That is, a pixel of differential image belongs to effective denoising pixel or invalid denoising
Pixel is predetermined, if the pixel is effective denoising pixel, then weight corresponding to the pixel is random true
Fixed, but it is random in default proportion range.
2nd, the ratio of the number of effective denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
And/or the position of invalid denoising pixel is random, weight corresponding to effective denoising pixel is random in default proportion range.
In this case, a pixel of differential image, which belongs to effective denoising pixel or invalid denoising pixel, is
Determine at random, corresponding weight is also to be determined at random in default proportion range, but is the need to ensure that effective denoising pixel
The ratio of the number of point and the number of invalid denoising pixel is fixed.
3rd, the ratio of the number of effective denoising pixel and the number of invalid denoising pixel is random, effective denoising pixel
And/or the position of invalid denoising pixel is random, weight corresponding to effective denoising pixel is fixed.
In this case, a pixel of differential image, which belongs to effective denoising pixel or invalid denoising pixel, is
Determine at random, if the pixel is effective denoising pixel, weight corresponding to the pixel is fixed, specific value
To preset some value in proportion range.
4th, the ratio of the number of effective denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
And/or the position of invalid denoising pixel is random, weight corresponding to effective denoising pixel is fixed.
In this case, a pixel of differential image, which belongs to effective denoising pixel or invalid denoising pixel, is
Determine at random, but need to ensure the ratio of the number of effective denoising pixel of differential image and the number of invalid denoising pixel
Example is fixed.If the pixel is effective denoising pixel, weight corresponding to the pixel is fixed, specific value
To preset some value in proportion range.
Those skilled in the art can select one of which to determine each pixel of differential image from above-mentioned four kinds of situations
Corresponding weight.
S103：It is right according to the pixel value of the differential image pixel and the value of the denoising parameter determined at random
The original image or the Pre-denoised image carry out denoising, obtain target image.
Weight corresponding to each pixel of the differential image is being determined according to the value of the denoising parameter determined at random
Afterwards, the pixel value according to the differential image pixel and corresponding weight, to the original image or the Pre-denoised image
Denoising is carried out, obtains target image.
Because the first parameter, the second parameter and the 3rd parameter wherein at least one determine at random, so target figure
It seem random image.In actual applications, multiple target images can be generated, then in response to from entering after multiple processing in image
The selection instruction of row selection, show image after selected processing.
Wherein, it can be that user performs to carry out the action of selection from multiple target images, i.e., enter multiple target images
Row display, selects optimum target image by user, can also be performed automatically by terminal, the foundation performed automatically can be to target
The judgement of image effect, i.e. terminal can pick out the target image of best results according to preset rules.Preset rules can be
The overall target signal to noise ratio of image is preset, then selects signal to noise ratio closest to the target image of the target signal to noise ratio.Target
Signal to noise ratio may be replaced by target image spatial resolution, target image density resolution, target image contrast etc..When
So, there can also be other preset rules, those skilled in the art can voluntarily select according to practical application.
In addition, the value of fixed object can be determined by user, can also be automatically determined by terminal.Determined when by user
When, user can be manually entered the value, can also design one " slide bar ", and the both ends of " slide bar " represent the model of the value of fixed object
Enclose, by sliding " sliding block " on " slide bar ", reach the purpose for the value for changing fixed object.In actual applications, change is passed through
The value of fixed object, can generate different target images, then can be chosen from multiple target images effect it is best one
It is individual.
Referring to Fig. 2 and Fig. 3, the multiple target images of Fig. 2 and Fig. 3 to be obtained according to above-mentioned second of situation.Fig. 2 and figure
Unlike 3, the noise removed in Fig. 2 is white noise, and the noise removed in Fig. 3 is motion artifacts.Wherein, each target figure
The second parameter (being represented in figs. 2 and 3 with r) is different as corresponding to, fixed corresponding to each target image in Fig. 2 and Fig. 3
The second parameter be respectively arranged to 0,10%, 30%, 50%, 70% and 100%.The determination of 3rd stochastic parameter, the first parameter
Default proportion range (0-1] in it is random.
According to Fig. 2 six width figures can be seen that with effective denoising pixel number and invalid denoising pixel number it
Between ratio gradual increase, the white noise carried in target image is more and more obvious；And with effective denoising pixel number and
Ratio is gradually reduced between invalid denoising pixel number, and the image information of target image is more and more unintelligible.Gone when effectively
When ratio is 30% between pixel number of making an uproar and invalid denoising pixel number, obtained target image effect is best.
It is can be seen that according to Fig. 3 six width figures when between effective denoising pixel number and invalid denoising pixel number
When ratio is 10%, obtained target image effect is best.
In addition, it is optional, if the noise information of original image could be got in advance, such as area-of-interest, noise would be strong
The information such as probability distribution are spent, then the noise information can also be considered during denoising, i.e., are believed according to the noise of the original image
Breath, the pixel value and weight of the differential image pixel, the original image or the Pre-denoised image are carried out at denoising
Reason.The noise information of original image i.e. the noise information of differential image, denoising is carried out according to the noise information of original image,
Also it is equivalent to carry out denoising according to the noise information of differential image.
, can be according to above-mentioned for the pixel in area-of-interest if noise information includes area-of-interest
Four kinds of situations are come weight corresponding to determining.And for the overseas pixel of region of interest, without exception as invalid denoising pixel
Point processing.
Noise intensity probability distribution embodies the relativeness of the noise intensity between each pixel.Noise intensity probability point
Cloth can obtain in the following way：
Mode one, it is being scanned to human body progress MRI (Magnetic Resonance Imaging, magnetic resonance imaging)
Before, Noise scan can be carried out.So-called Noise scan refers to magnetic resonance excitation to close, and only opens and receives.Due to activating system
Close, so being produced without signal.At this moment the only noise signal that system receives, so as to obtain noise intensity probability distribution.
Mode two, parallel imaging.The geometrical factor g-factor of parallel imaging is the Theoretical Prediction to noise profile, therefore
Noise intensity probability distribution can be obtained according to the g-factor of parallel imaging.
Certainly, above two mode is not the restriction to obtaining noise intensity probability distribution, and those skilled in the art are also
It can be obtained using other means.
By considering noise prior information during denoising so that denoising process is more effective, obtained target figure
Picture it is better.
The present embodiment determines denoising parameter corresponding to the differential image pixel at random by obtaining differential image
Value, then according to the pixel value of the differential image pixel and the value of the denoising parameter determined at random, to the original
Image or the Pre-denoised image carry out denoising, obtain target image.If the target image generated at random is not
The image that user intentionally gets, then the target image of multiple different noise levels, Cong Zhongxuan can be generated by adjusting " slide bar "
The target image of a best results is selected, realizes that balances noise removes and retained the purpose of image information.
The image processing method that the present embodiment provides can both be applied in the time domain, can also apply in a frequency domain.Below
It is introduced respectively with reference to specific practical application scene.
The technical scheme in frequency domain application is introduced first.
Referring to Fig. 4, the figure is a kind of flow chart for image processing method that the embodiment of the present application provides.
The image processing method that the present embodiment provides comprises the following steps：
S201：Time domain original image I is obtained, and determines predenoising parameter value.
The time domain original image is the original image that picture signal is time-domain signal.
S202：Predenoising is carried out to time domain original image according to predenoising parameter value, obtains time domain Pre-denoised image
Time domain Pre-denoised image is the Pre-denoised image that picture signal is time-domain signal.
S203：The time domain original image I is subjected to Fourier transformation, obtains frequency domain original image k.
The frequency domain original image k is the original image that picture signal is frequency-region signal.
S204：By the time domain Pre-denoised imageFourier transformation is carried out, obtains frequency domain Pre-denoised image
The frequency domain Pre-denoised imageIt is the Pre-denoised image that picture signal is frequency-region signal.
S205：By the frequency domain original image k and the frequency domain Pre-denoised imageSubtract each other, obtain differential image
S206：Time domain original image I noise information is obtained, and differential image Δ k is optimized according to noise information, is optimized
Differential image Δ k' afterwards.
Assuming that noise information can be represented with matrix N, each picture with differential image Δ k respectively of the element in matrix N
Vegetarian refreshments corresponds, the so-called one-to-one correspondence being meant that on position, for example, in matrix N the first row first row element
Corresponding to the pixel of differential image Δ k the first row first rows.The value of element is optimization numerical value in matrix N.If for example, noise
Information is area-of-interest, then the value of element corresponding with area-of-interest can be 1 in matrix N, remaining element, that is, feel
The value of element corresponding to pixel outside interest region can be 0.For another example if noise information is noise intensity probability point
Cloth, then the value of each element is the noise intensity probable value of corresponding pixel points in matrix N.
In actual applications, differential image Δ k, that is, the differential image after optimizing can be optimized according to equation below
S207：The value of denoising parameter corresponding to the differential image Δ k' pixels is determined at random.
S208：Join according to the pixel value of differential image Δ k' pixels after the optimization and the denoising determined at random
Several value, to the frequency domain original image k or described frequency domain Pre-denoised imagesCarry out denoising.
Definition matrix M, M size are identical with the size of matrix N, i.e. the number of element is identical.Each element in matrix M
Respectively each pixel with differential image is corresponded, and one-to-one implication is same as above.The value of element is difference in matrix M
The weight of image corresponding pixel points, weight can be random, can also fix, referring specifically to four kinds of random cases above.
Target image K can be calculated according to equation below：
Or k=k-M Δs k
Carry out denoising reason in domain space to be, in the differential image k to domain space~Some pixel weighting
When weight, it is split in time domain equivalent to weight on all pixels point, so the smoothness of image is higher, effect is preferable.
S209：Target image K is subjected to Fourier inversion, obtains the target image in time domain space.
Because the image generally shown is all time-domain signal, need target image K by Fourier inversion
Corresponding signal is converted to time domain by frequency domain.
The technical scheme in time domain application is described below.
Referring to Fig. 5, the figure is a kind of flow chart for image processing method that the embodiment of the present application provides.
The image processing method that the present embodiment provides comprises the following steps：
S301：Time domain original image I is obtained, and determines predenoising parameter value.
The time domain original image is the original image that picture signal is time-domain signal.
S302：Predenoising is carried out to time domain original image I according to predenoising parameter value, obtains time domain Pre-denoised image
Time domain Pre-denoised image is the Pre-denoised image that picture signal is time-domain signal.
S303：By time domain original image I and time domain Pre-denoised imageSubtract each other, obtain differential image Δ I.
S304：Time domain original image I noise information is obtained, and differential image Δ I is optimized according to noise information, is optimized
Differential image afterwards
Wherein matrix N represents time domain original image I noise information, and the element in matrix N is each with time domain original image I respectively
Individual pixel corresponds.
S305：The value of denoising parameter corresponding to the differential image Δ I' pixels is determined at random.
S306：Join according to the pixel value of differential image Δ I' pixels after the optimization and the denoising determined at random
Several value, to time domain original image I or time domain Pre-denoised imageDenoising is carried out, obtains target image S.
Definition matrix M, M size are identical with the size of matrix N, i.e. the number of element is identical.Each element in matrix M
Respectively corresponded with each pixel of differential image or original image, one-to-one implication is same as above.Element in matrix M
Being worth can be random for the weight of differential image Δ I' pixels after optimization, weight, can also fix, referring specifically to four above
Kind random case.
Target image S can be calculated according to equation below：
Or S=I-M Δs I'
Because step S301 to S306 is similar with S201 to S208, the specific descriptions of identical part referring to S201 extremely
S208。
It should be noted that the execution sequence of step does not form the restriction to the application in above-described embodiment, can
On the premise of realizing the present application purpose, those skilled in the art can be according to actual conditions voluntarily reversed order.
A kind of image processing method provided based on above example, the embodiment of the present application additionally provide a kind of image procossing
Device, describe its operation principle in detail below in conjunction with the accompanying drawings.
Device embodiment
Referring to Fig. 6, the figure is a kind of structured flowchart for image processing apparatus that the embodiment of the present application provides.
The image processing apparatus that the present embodiment provides includes：
Differential image acquiring unit 101, determining unit 102 and denoising unit 103；
Wherein, the differential image acquiring unit 101, for obtaining differential image, the differential image be original image and
Pre-denoised image subtracts each other obtained image, and the Pre-denoised image is image of the original image after predenoising, Mei Gesuo
The pixel for stating differential image is all corresponding with weight respectively, and the pixel of the differential image includes effective denoising pixel and nothing
Imitate denoising pixel, the effectively denoising pixel be corresponding to weight be not default weight pixel, the invalid denoising
Pixel is the pixel that corresponding weight is default weight；
The determining unit 102, it is described for determining the value of denoising parameter corresponding to the differential image pixel at random
Denoising parameter includes the first parameter, and/or the second parameter, and/or the 3rd parameter；Wherein, first parameter is in default power
The weight effectively corresponding to denoising pixel in weight scope, second parameter for the effectively number of denoising pixel and
Ratio between the number of invalid denoising pixel, the 3rd parameter are the effectively denoising pixel and/or described invalid
The position of denoising pixel；
The denoising unit 103, determine for the pixel value according to the differential image pixel and at random described in
The value of denoising parameter, denoising is carried out to the original image or the Pre-denoised image.
The present embodiment is being preset in proportion range to corresponding to the effectively denoising pixel by obtaining differential image
Weight determined at random, and/or, the value for presetting denoising parameter is determined at random, finally according to the differential image picture
The pixel value and weight of vegetarian refreshments, denoising is carried out to the original image or the Pre-denoised image, obtains target image.Such as
The target image that fruit generates at random is not the image that user intentionally gets, then can generate multiple differences by adjusting slide bar
The target image of noise level, the target image of a best results is therefrom selected, realize that balances noise removes and retained image
The purpose of information.
Optionally, the denoising unit 103 includes：
Noise information acquiring unit and denoising subelement；
The noise information acquiring unit, for obtaining the noise information of the original image；
The denoising subelement, for optimizing the differential image according to the noise information of the original image.
Optionally, the noise information includes following at least one：
Area-of-interest and noise intensity probability distribution.
Optionally, the differential image acquiring unit includes：
It is single that time-domain image acquiring unit, original image converter unit, Pre-denoised image converter unit and differential image obtain son
Member；
Wherein, the time-domain image acquiring unit, for obtaining time domain original image and time domain Pre-denoised image, the time domain
Original image is the original image that picture signal is time-domain signal, and time domain Pre-denoised image is the predenoising that picture signal is time-domain signal
Image；
The original image converter unit, for the time domain original image to be carried out into Fourier transformation, frequency domain original image is obtained,
The frequency domain original image is the original image that picture signal is frequency-region signal；
The Pre-denoised image converter unit, for the time domain Pre-denoised image to be carried out into Fourier transformation, obtain frequency
Domain Pre-denoised image, the frequency domain Pre-denoised image are the Pre-denoised images that picture signal is frequency-region signal；
The differential image obtains subelement, for the frequency domain original image and the frequency domain Pre-denoised image to be subtracted each other,
Obtain the differential image；
The denoising unit, is specifically used for：
According to the pixel value and weight of the differential image pixel, the frequency domain original image or the frequency domain are gone in advance
Image of making an uproar carries out denoising.
Optionally, the determining unit, is specifically used for：
The effectively ratio of the number of denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
The position of point and/or invalid denoising pixel is fixed, and weight corresponding to effective denoising pixel is random in default proportion range；
Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
And/or the position of invalid denoising pixel is random, weight corresponding to effective denoising pixel is random in default proportion range；Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is random, effective denoising pixel
And/or the position of invalid denoising pixel is random, weight corresponding to effective denoising pixel is fixed；Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
And/or the position of invalid denoising pixel is random, weight corresponding to effective denoising pixel is fixed.
The image processing apparatus that the present embodiment provides can be applied on any electronic equipment with processor, the electricity
Sub- equipment can be existing, researching and developing or research and development in the future any electronic equipments, include but is not limited to：It is existing, just
Research and development or in the future research and development desktop computers, laptop computer, mobile terminal (including smart mobile phone, non-smart mobile phone,
Various tablet personal computers) etc..Device embodiment can be realized by software, can also be by way of hardware or software and hardware combining
Realize.It is the electronic equipment by carrying processor where it as the device on a logical meaning exemplified by implemented in software
Processor corresponding computer program instructions in memory are read in internal memory what operation was formed.For hardware view,
As shown in fig. 7, a kind of hardware structure diagram of the electronic equipment of processor is carried where the application image processing apparatus, except figure
Outside processor, internal memory, network interface and memory shown in 7, the electronics with processor in embodiment where device
Equipment can also include other hardware, such as display, this is repeated no more generally according to the actual functional capability of the equipment.
Wherein, logical order corresponding to image processing method can be stored with memory, the memory for example can be
Nonvolatile memory (non-volatile memory), processor can call the logic for performing the preservation in memory to refer to
Order, to perform above-mentioned image processing method, display is used to show the image after carrying out denoising.
The function of logical order corresponding to image processing method, if realized in the form of software function module and as solely
Vertical production marketing in use, can be stored in a computer read/write memory medium.Based on such understanding, this hair
The part or the part of the technical scheme that bright technical scheme substantially contributes to prior art in other words can be with soft
The form of part product is embodied, and the computer software product is stored in a storage medium, including some instructions are making
Obtain a computer equipment (can be personal computer, server, or network equipment etc.) and perform each embodiment of the present invention
The all or part of step of methods described.And foregoing storage medium includes：USB flash disk, mobile hard disk, read-only storage (ROM,
Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disc or CD etc. it is various
Can be with the medium of store program codes.
A kind of image processing method and device provided based on above example, the embodiment of the present application additionally provide a kind of figure
As processing equipment, its operation principle is described in detail below in conjunction with the accompanying drawings.
Apparatus embodiments
This application provides a kind of image processing equipment, the equipment includes：Processor, can for storing the processor
The memory and display of execute instruction；
Wherein, the processor is configured as：
Differential image is obtained, the differential image is the image that original image and Pre-denoised image subtract each other to obtain, described pre- to go
Image of making an uproar is image of the original image after predenoising, and the pixel of each differential image corresponds to respectively has the right
Weight, the pixel of the differential image include effective denoising pixel and invalid denoising pixel, the effectively denoising pixel
It is not the pixel of default weight for corresponding weight, the invalid denoising pixel is the picture that corresponding weight is default weight
Vegetarian refreshments；
The value of denoising parameter corresponding to the differential image pixel is determined at random, and the denoising parameter includes the first ginseng
Number, and/or the second parameter, and/or the 3rd parameter；Wherein, first parameter is effectively gone to be described in default proportion range
Make an uproar weight corresponding to pixel, second parameter is the effectively number of denoising pixel and of invalid denoising pixel
Ratio between number, the 3rd parameter for effectively denoising pixel and/or the invalid denoising pixel position according to
According to the pixel value of the differential image pixel and the value of the denoising parameter determined at random, to the original image or described
Pre-denoised image carries out denoising；
The display, for showing the image after carrying out denoising.
When introducing the element of various embodiments of the application, article "a", "an", "this" and " described " are intended to
Indicate one or more elements.Word " comprising ", "comprising" and " having " are all inclusive and meaned except listing
Outside element, there can also be other elements.
It should be noted that one of ordinary skill in the art will appreciate that realize the whole in above method embodiment or portion
Split flow, it is that by computer program the hardware of correlation can be instructed to complete, described program can be stored in a computer
In read/write memory medium, the program is upon execution, it may include such as the flow of above-mentioned each method embodiment.Wherein, the storage
Medium can be magnetic disc, CD, read-only memory (Read-Only Memory, ROM) or random access memory (Random
Access Memory, RAM) etc..
Each embodiment in this specification is described by the way of progressive, identical similar portion between each embodiment
Divide mutually referring to what each embodiment stressed is the difference with other embodiment.It is real especially for device
For applying example, because it is substantially similar to embodiment of the method, so describing fairly simple, related part is referring to embodiment of the method
Part explanation.Device embodiment described above is only schematical, wherein described be used as separating component explanation
Unit and module can be or may not be physically separate.Furthermore it is also possible to it is selected according to the actual needs
In some or all of unit and module realize the purpose of this embodiment scheme.Those of ordinary skill in the art are not paying
In the case of creative work, you can to understand and implement.
Described above is only the embodiment of the application, it is noted that for the ordinary skill people of the art
For member, on the premise of the application principle is not departed from, some improvements and modifications can also be made, these improvements and modifications also should
It is considered as the protection domain of the application.
Claims (10)
1. a kind of image processing method, it is characterised in that methods described includes：
Obtain differential image, the differential image is the image that original image and Pre-denoised image subtract each other to obtain, the predenoising figure
As being image of the original image after predenoising, the pixel of each differential image is corresponding with weight respectively, institute
Stating the pixel of differential image includes effective denoising pixel and invalid denoising pixel, and the effectively denoising pixel is corresponding
Weight be not the pixel of default weight, the invalid denoising pixel be corresponding to weight be default weight pixel；
Determine the value of denoising parameter corresponding to the differential image pixel at random, the denoising parameter include the first parameter and/
Or second parameter, and/or the 3rd parameter；Wherein, first parameter is the effectively denoising pixel in default proportion range
Weight corresponding to point, second parameter is between the effectively number of denoising pixel and the number of invalid denoising pixel
Ratio, the 3rd parameter for effectively denoising pixel and/or the invalid denoising pixel position；
According to the pixel value of the differential image pixel and the value of the denoising parameter determined at random, to the original image
Or the Pre-denoised image carries out denoising, obtains target image.
2. according to the method for claim 1, it is characterised in that this method also includes：
Obtain the noise information of the original image；
The differential image is optimized according to the noise information of the original image.
3. according to the method for claim 2, it is characterised in that the noise information includes following at least one：
Area-of-interest and noise intensity probability distribution.
4. according to the method described in claims 1 to 3 any one, it is characterised in that the acquisition differential image includes：
Time domain original image and time domain Pre-denoised image are obtained, the time domain original image is the artwork that picture signal is time-domain signal
Picture, time domain Pre-denoised image are the Pre-denoised images that picture signal is time-domain signal；
The time domain original image is subjected to Fourier transformation, obtains frequency domain original image, the frequency domain original image is that picture signal is
The original image of frequency-region signal；
The time domain Pre-denoised image is subjected to Fourier transformation, obtains frequency domain Pre-denoised image, the frequency domain Pre-denoised image
It is the Pre-denoised image that picture signal is frequency-region signal；
The frequency domain original image and the frequency domain Pre-denoised image are subtracted each other, obtain the differential image；
It is described that the original image or Pre-denoised image progress denoising are included：
Denoising is carried out to the frequency domain original image or the frequency domain Pre-denoised image.
5. according to the method for claim 1, it is characterised in that described to determine at random corresponding to the differential image pixel
The value of denoising parameter includes：
The effectively ratio of the number of denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
And/or the position of invalid denoising pixel is fixed, weight corresponding to effective denoising pixel is random in default proportion range；Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel and/or
The position of invalid denoising pixel is random, and weight corresponding to effective denoising pixel is random in default proportion range；Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is random, effective denoising pixel and/or
The position of invalid denoising pixel is random, and weight corresponding to effective denoising pixel is fixed；Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel and/or
The position of invalid denoising pixel is random, and weight corresponding to effective denoising pixel is fixed.
6. a kind of image processing apparatus, it is characterised in that described device includes：
Differential image acquiring unit, determining unit and denoising unit；
Wherein, the differential image acquiring unit, for obtaining differential image, the differential image is original image and predenoising figure
As subtracting each other obtained image, the Pre-denoised image is image of the original image after predenoising, each disparity map
The pixel of picture is all corresponding with weight respectively, and the pixel of the differential image includes effective denoising pixel and invalid denoising picture
Vegetarian refreshments, the effectively denoising pixel are the pixel that corresponding weight is not default weight, and the invalid denoising pixel is
Corresponding weight is the pixel of default weight；
The determining unit, for determining the value of denoising parameter corresponding to the differential image pixel, the denoising ginseng at random
Number includes the first parameter, and/or the second parameter, and/or the 3rd parameter；Wherein, first parameter is in default proportion range
The interior weight effectively corresponding to denoising pixel, second parameter is the effectively number of denoising pixel and invalid goes
Ratio between the number of pixel of making an uproar, the 3rd parameter are effectively denoising pixel and/or the invalid denoising picture
The position of vegetarian refreshments；
The denoising unit, for the pixel value according to the differential image pixel and the denoising parameter determined at random
Value, denoising is carried out to the original image or the Pre-denoised image, obtains target image.
7. device according to claim 6, it is characterised in that described device also includes：
Noise information acquiring unit and denoising subelement；
The noise information acquiring unit, for obtaining the noise information of the original image；
The denoising subelement, for optimizing the differential image according to the noise information of the original image.
8. according to the device described in claim 6-7 any one, it is characterised in that the differential image acquiring unit includes：
Time-domain image acquiring unit, original image converter unit, Pre-denoised image converter unit and differential image obtain subelement；
Wherein, the time-domain image acquiring unit, for obtaining time domain original image and time domain Pre-denoised image, the time domain artwork
Seem the original image that picture signal is time-domain signal, time domain Pre-denoised image is the predenoising figure that picture signal is time-domain signal
Picture；
The original image converter unit, for the time domain original image to be carried out into Fourier transformation, frequency domain original image is obtained, it is described
Frequency domain original image is the original image that picture signal is frequency-region signal；
The Pre-denoised image converter unit, for the time domain Pre-denoised image to be carried out into Fourier transformation, it is pre- to obtain frequency domain
Denoising image, the frequency domain Pre-denoised image are the Pre-denoised images that picture signal is frequency-region signal；
The differential image obtains subelement, for the frequency domain original image and the frequency domain Pre-denoised image to be subtracted each other, obtains
The differential image；
The denoising unit, is specifically used for：
According to the pixel value and weight of the differential image pixel, to the frequency domain original image or the frequency domain predenoising figure
As carrying out denoising, target image is obtained.
9. device according to claim 6, it is characterised in that the determining unit, be specifically used for：
The effectively ratio of the number of denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel
And/or the position of invalid denoising pixel is fixed, weight corresponding to effective denoising pixel is random in default proportion range；Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel and/or
The position of invalid denoising pixel is random, and weight corresponding to effective denoising pixel is random in default proportion range；Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is random, effective denoising pixel and/or
The position of invalid denoising pixel is random, and weight corresponding to effective denoising pixel is fixed；Or
The ratio of the number of effective denoising pixel and the number of invalid denoising pixel is fixed, effective denoising pixel and/or
The position of invalid denoising pixel is random, and weight corresponding to effective denoising pixel is fixed.
10. a kind of image processing equipment, it is characterised in that the equipment includes：Processor, it can hold for storing the processor
The memory and display of row instruction；
Wherein, the processor is configured as：
Obtain differential image, the differential image is the image that original image and Pre-denoised image subtract each other to obtain, the predenoising figure
As being image of the original image after predenoising, the pixel of each differential image is corresponding with weight respectively, institute
Stating the pixel of differential image includes effective denoising pixel and invalid denoising pixel, and the effectively denoising pixel is corresponding
Weight be not the pixel of default weight, the invalid denoising pixel be corresponding to weight be default weight pixel；
Determine the value of denoising parameter corresponding to the differential image pixel at random, the denoising parameter include the first parameter and/
Or second parameter, and/or the 3rd parameter；Wherein, first parameter is the effectively denoising pixel in default proportion range
Weight corresponding to point, second parameter is between the effectively number of denoising pixel and the number of invalid denoising pixel
Ratio, the 3rd parameter for effectively denoising pixel and/or the invalid denoising pixel position；
According to the pixel value of the differential image pixel and the value of the denoising parameter determined at random, to the original image
Or the Pre-denoised image carries out denoising, obtains target image；
The display, for showing the target image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710666951.6A CN107451978B (en) | 2017-08-07 | 2017-08-07 | Image processing method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710666951.6A CN107451978B (en) | 2017-08-07 | 2017-08-07 | Image processing method, device and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107451978A true CN107451978A (en) | 2017-12-08 |
| CN107451978B CN107451978B (en) | 2020-08-21 |
Family
ID=60490929
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710666951.6A Active CN107451978B (en) | 2017-08-07 | 2017-08-07 | Image processing method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107451978B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108492246A (en) * | 2018-03-12 | 2018-09-04 | 维沃移动通信有限公司 | A kind of image processing method, device and mobile terminal |
| CN113781357A (en) * | 2021-09-24 | 2021-12-10 | Oppo广东移动通信有限公司 | Image processing method, image processing device, electronic equipment and storage medium |
| US20220036518A1 (en) * | 2020-07-31 | 2022-02-03 | Beijing Xiaomi Mobile Software Co., Ltd. | Method for processing image, electronic device and storage medium |
| CN115861462A (en) * | 2022-10-17 | 2023-03-28 | 北京百度网讯科技有限公司 | Training method and device for image generation model, electronic equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104680536A (en) * | 2015-03-09 | 2015-06-03 | 西安电子科技大学 | Method for detecting SAR image change by utilizing improved non-local average algorithm |
| CN104751416A (en) * | 2013-12-31 | 2015-07-01 | 展讯通信（上海）有限公司 | Image denoising method and device and image processing system |
| CN105844599A (en) * | 2016-04-11 | 2016-08-10 | 沈阳东软医疗系统有限公司 | De-noising method and apparatus for image reconstruction |
| CN106355561A (en) * | 2016-08-30 | 2017-01-25 | 天津大学 | TV (total variation) image noise removal method based on noise priori constraint |
| CN106780352A (en) * | 2016-12-16 | 2017-05-31 | 珠海赛纳打印科技股份有限公司 | Image rotating method, device and image forming apparatus |
| CN106875354A (en) * | 2017-01-20 | 2017-06-20 | 北京东软医疗设备有限公司 | Image de-noising method, device and equipment |
-
2017
- 2017-08-07 CN CN201710666951.6A patent/CN107451978B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104751416A (en) * | 2013-12-31 | 2015-07-01 | 展讯通信（上海）有限公司 | Image denoising method and device and image processing system |
| CN104680536A (en) * | 2015-03-09 | 2015-06-03 | 西安电子科技大学 | Method for detecting SAR image change by utilizing improved non-local average algorithm |
| CN105844599A (en) * | 2016-04-11 | 2016-08-10 | 沈阳东软医疗系统有限公司 | De-noising method and apparatus for image reconstruction |
| CN106355561A (en) * | 2016-08-30 | 2017-01-25 | 天津大学 | TV (total variation) image noise removal method based on noise priori constraint |
| CN106780352A (en) * | 2016-12-16 | 2017-05-31 | 珠海赛纳打印科技股份有限公司 | Image rotating method, device and image forming apparatus |
| CN106875354A (en) * | 2017-01-20 | 2017-06-20 | 北京东软医疗设备有限公司 | Image de-noising method, device and equipment |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108492246A (en) * | 2018-03-12 | 2018-09-04 | 维沃移动通信有限公司 | A kind of image processing method, device and mobile terminal |
| US20220036518A1 (en) * | 2020-07-31 | 2022-02-03 | Beijing Xiaomi Mobile Software Co., Ltd. | Method for processing image, electronic device and storage medium |
| US11756167B2 (en) * | 2020-07-31 | 2023-09-12 | Beijing Xiaomi Mobile Software Co., Ltd. | Method for processing image, electronic device and storage medium |
| CN113781357A (en) * | 2021-09-24 | 2021-12-10 | Oppo广东移动通信有限公司 | Image processing method, image processing device, electronic equipment and storage medium |
| CN115861462A (en) * | 2022-10-17 | 2023-03-28 | 北京百度网讯科技有限公司 | Training method and device for image generation model, electronic equipment and storage medium |
| CN115861462B (en) * | 2022-10-17 | 2023-11-03 | 北京百度网讯科技有限公司 | Training method and device for image generation model, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107451978B (en) | 2020-08-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107451978A (en) | A kind of image processing method, device and equipment | |
| CN111369440B (en) | Model training and image super-resolution processing method, device, terminal and storage medium | |
| Saladi et al. | Analysis of denoising filters on MRI brain images | |
| TWI260158B (en) | Electronic watermark embedding device, electronic watermark detection device, method thereof, and program | |
| CN110717857A (en) | Super-resolution image reconstruction method and device | |
| US20110274329A1 (en) | Processing of medical image data | |
| US9292910B2 (en) | Image processing apparatus, image processing method, program, printing medium, and recording medium | |
| CN113269858B (en) | Virtual scene rendering method and device, computer equipment and storage medium | |
| Kovesi | Phase preserving tone mapping of non-photographic high dynamic range images | |
| US20160300331A1 (en) | Scalable massive parallelization of overlapping patch aggregation | |
| CN110059666A (en) | A kind of attention detection method and device | |
| CN110910326B (en) | Image processing method and device, processor, electronic equipment and storage medium | |
| CN111091517B (en) | Residual weighted imaging method and device | |
| Wang et al. | A novel attention-guided JND Model for improving robust image watermarking | |
| JP4993146B2 (en) | System, method and computer program for preventing the display of unwanted content stored in a frame buffer | |
| Chen et al. | An evolutionary game-theoretic approach for image interpolation | |
| Cordes et al. | Optimizing the performance of local canonical correlation analysis in fMRI using spatial constraints | |
| CN112116102A (en) | Method and system for expanding domain adaptive training set | |
| Dong et al. | Exploiting entropy masking in perceptual graphic rendering | |
| Hung et al. | Moran’s I for impulse noise detection and removal in color images | |
| WO2011086594A1 (en) | Image processing apparatus and method therefor | |
| DE102011013584A1 (en) | Hardware-accelerated simulation of atmospheric scattering | |
| JP5447867B2 (en) | Image restoration apparatus and image restoration method | |
| Annighofer et al. | Prediction of results from subjective evaluation of real-time-capable tone-mapping operators applied to limited high-dynamic-range images | |
| Ryu et al. | Improving high frequency image features of deep learning reconstructions via k‐space refinement with null‐space kernel |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |