CN109584172A - Backlight compensation method and device based on the fuzzy learning machine that transfinites of iteration - Google Patents
Backlight compensation method and device based on the fuzzy learning machine that transfinites of iteration Download PDFInfo
- Publication number
- CN109584172A CN109584172A CN201811337780.3A CN201811337780A CN109584172A CN 109584172 A CN109584172 A CN 109584172A CN 201811337780 A CN201811337780 A CN 201811337780A CN 109584172 A CN109584172 A CN 109584172A
- Authority
- CN
- China
- Prior art keywords
- backlight
- degree
- compensated
- transfinites
- learning machine
- 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.)
- Withdrawn
Links
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000012549 training Methods 0.000 claims description 23
- 230000003044 adaptive effect Effects 0.000 claims description 16
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000004422 calculation algorithm Methods 0.000 claims description 9
- 230000001186 cumulative effect Effects 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 6
- 238000012546 transfer Methods 0.000 claims description 6
- 230000001052 transient effect Effects 0.000 claims description 5
- 238000004891 communication Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 235000008331 Pinus X rigitaeda Nutrition 0.000 description 1
- 235000011613 Pinus brutia Nutrition 0.000 description 1
- 241000018646 Pinus brutia Species 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 210000003746 feather Anatomy 0.000 description 1
- 244000144992 flock Species 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000005070 sampling 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
- G06T5/90—Dynamic range modification of images or parts thereof
-
- 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
-
- 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
- G06T2207/20081—Training; Learning
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Circuit Arrangement For Electric Light Sources In General (AREA)
Abstract
The embodiment of the present invention provides a kind of backlight compensation method and device based on the fuzzy learning machine that transfinites of iteration, the described method includes: the first backlight degree and the second backlight degree of the backlight image to be compensated that will acquire, it is input to the fuzzy learning machine that transfinites of trained iteration, exports third backlight degree；The backlight image to be compensated is compensated based on the third backlight degree.Backlight compensation method and device provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration, two backlight degree indexs of backlight image to be compensated are returned by using the fuzzy learning machine that transfinites of iteration, obtain the final backlight degree of backlight image to be compensated, backlight image to be compensated is compensated based on final backlight degree, to improve the brightness of backlight image, the arithmetic speed for improving compensation expands the application range of backlight compensation.
Description
Technical field
The present embodiments relate to technical field of image processing more particularly to a kind of based on the fuzzy learning machine that transfinites of iteration
Backlight compensation method and device.
Background technique
With the development of science and technology the function of picture pick-up device is improved day by day, performance is also increasingly promoted.Although most of consumption
All there are many very powerful functions, such as auto-focusing, automatic exposure etc. for camera, and user can be allowed light under various shooting conditions
Pine shoots the photo of high quality, but user was it is possible to photographed backlight image, especially by monitoring camera, driving recording
The fixed focal lengths photographing device such as instrument.When master's (prospect) object and larger background luminance difference, backlight image will be generated.
In the prior art, there are many different methods to detect and compensate backlight image.Usually using the sky around object
Between information and related pixel statistical information, such as histogram equalization (HE) and histogram specification (HS), and using global straight
Side's figure equalization (GHE), partial histogram equalization (LHE) are based on conventional histogram equalization and dynamic histogram equalization
(DHE) improved models such as intelligent comparison degree Enhancement Method solve the problems, such as backlight.But these methods cannot concentrate on backlight
It region cannot further very because histogram information is only extracted by the way that histogram equalization is applied to luminance frequency distribution
Handle backlight situation well to prevent the interference to no backlight area.
In addition, by introducing fuzzy c-means, neural network, quality fuzzy rule and pixel-by-pixel backoff algorithm equal part is birdsed of the same feather flock together
Class technology divides the image into backlight area and non-backlight area.In Neuro-Fuzzy Network (FNFN) based on function link, according to
The deduction offset of FNFN generates compensated curve, and the method achieves preferable compensation effect.But this method exists to calculating
The high problem of Capability Requirement can not be applied in real-time scene.
Summary of the invention
A kind of overcome the above problem the purpose of the embodiment of the present invention is that providing or at least be partially solved the above problem
Backlight compensation method and device based on the fuzzy learning machine that transfinites of iteration.
In order to solve the above-mentioned technical problem, on the one hand, the embodiment of the present invention provides a kind of based on the fuzzy study of transfiniting of iteration
The backlight compensation method of machine, comprising:
The first backlight degree and the second backlight degree for the backlight image to be compensated that will acquire, are input to trained iteration mould
The learning machine that transfinites is pasted, third backlight degree is exported；
The backlight image to be compensated is compensated based on the third backlight degree.
On the other hand, the embodiment of the present invention provides a kind of backlight compensation device based on the fuzzy learning machine that transfinites of iteration,
It is characterized in that, comprising:
Computing module, the first backlight degree and the second backlight degree of the backlight image to be compensated for will acquire, is input to
The fuzzy learning machine that transfinites of trained iteration, exports third backlight degree；
Compensating module, for being compensated based on the third backlight degree to the backlight image to be compensated.
In another aspect, the embodiment of the present invention provides a kind of electronic equipment, comprising:
Memory and processor, the processor and the memory complete mutual communication by bus；It is described to deposit
Reservoir is stored with the program instruction that can be executed by the processor, and it is above-mentioned that the processor calls described program instruction to be able to carry out
Method.
Another aspect, the embodiment of the present invention provide a kind of non-transient computer readable storage medium, are stored thereon with calculating
Machine program realizes above-mentioned method when the computer program is executed by processor.
Backlight compensation method and device provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration, by using
The fuzzy learning machine that transfinites of iteration returns two backlight degree indexs of backlight image to be compensated, obtains backlight image to be compensated
Final backlight degree, backlight image to be compensated is compensated based on final backlight degree, to improve the brightness of backlight image,
The arithmetic speed for improving compensation expands the application range of backlight compensation.
Detailed description of the invention
Fig. 1 is the backlight compensation method schematic diagram provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration；
Fig. 2 is the backlight compensation schematic device provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration；
Fig. 3 is the structural schematic diagram of electronic equipment provided in an embodiment of the present invention.
Specific embodiment
In order to keep the purposes, technical schemes and advantages of the embodiment of the present invention clearer, implement below in conjunction with the present invention
Attached drawing in example, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment
It is a part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiment of the present invention, those of ordinary skill in the art
Every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
Fig. 1 is the backlight compensation method schematic diagram provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration, such as
Shown in Fig. 1, the embodiment of the present invention provides a kind of backlight compensation method based on the fuzzy learning machine that transfinites of iteration, and executing subject is
Based on the backlight compensation device (hereinafter referred to as compensation device) of the fuzzy learning machine that transfinites of iteration, this method comprises:
Step S101, the first backlight degree and the second backlight degree for the backlight image to be compensated that will acquire, are input to training
The fuzzy learning machine that transfinites of good iteration, exports third backlight degree；
Step S102, the backlight image to be compensated is compensated based on the third backlight degree.
Specifically, the backlight compensation method provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration includes inspection
Survey stage and compensated stage two parts.
Firstly, reading original backlight image to be compensated.
In detection-phase, the first backlight degree and the second backlight degree of backlight image to be compensated are obtained, backlight degree is for indicating
The backlight degree of backlight image, and the degree of light filling is needed, the value of backlight degree reflects more greatly the backlight degree of backlight image
It is bigger, need the degree of light filling also bigger.First backlight degree and the second backlight degree are the backlights to be compensated that primary Calculation obtains
The backlight degree of image.
Then, the first backlight degree and the second backlight degree are input to the fuzzy learning machine that transfinites of trained iteration, output the
Three backlight degrees.That is, by using the fuzzy learning machine that transfinites of iteration to the first backlight degree and the second backlight of backlight image to be compensated
It is returned, obtains third backlight degree, which is the backlight degree of final, accurate backlight image to be compensated.Repeatedly
The generation fuzzy learning machine that transfinites is to transfinite what learning machine improved to fuzzy using Boosting method.By to fuzzy super
The improvement for limiting learning machine improves the robustness of the fuzzy learning machine that transfinites.
Finally, being compensated based on third backlight degree to backlight image to be compensated.To improve the part of backlight image
Contrast remains the details of image, and image is made to seem more natural, and can facilitate under various ambient lighting conditions and angle
Effectively enhance image, and improves compensation speed.
Backlight compensation method provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration, by using iteration mould
It pastes the learning machine that transfinites to return two backlight degree indexs of backlight image to be compensated, obtains the final of backlight image to be compensated
Backlight degree compensates backlight image to be compensated based on final backlight degree, to improve the brightness of backlight image, improves
The arithmetic speed of compensation expands the application range of backlight compensation.
On the basis of the above embodiments, further, the training method of the fuzzy learning machine that transfinites of the iteration is as follows:
Obtain the preset stopping condition and default the number of iterations of the fuzzy learning machine that transfinites of training；
Several training samples that will have been marked in advance are input to the fuzzy learning machine that transfinites and are iterated training, work as iteration
When number meets the preset stopping condition when reaching the default the number of iterations or certain iteration, then training terminates, and obtains
The fuzzy learning machine that transfinites of trained iteration.
Specifically, before being compensated to backlight image to be compensated, need it is first fuzzy to iteration transfinite learning machine into
Row training.Training iteration is fuzzy to transfinite that the specific method is as follows for learning machine:
Firstly, the value of the stop condition stop of the fuzzy learning machine that transfinites of configuration training is denoted as the value of S and the number of iterations m
It is denoted as M, putting back to sample rate is (0 < α≤1) α.Each sample standard deviation is the different backlight image of backlight degree, each training image sample
This all has already passed through calculating, is marked with two backlight factor indexs, that is, the first backlight degree and the second backlight degree, i-th of image
Two backlight factor indexs be denoted as x with vectori, it is also marked with third backlight degree, i.e., final, accurate backlight degree is denoted as
yi, then being denoted as N number of training image sample
Then, this N number of training image sample that will have been marked in advance is input to the fuzzy learning machine that transfinites and is iterated instruction
Practice.
(1) using the fuzzy learning machine algorithm that transfinites, gained training set is sampled with putting back toTraining is fuzzy super
Limit learning machine network, available model f0And F (x),0(x)=f0(x)。
(2) for m=1 to M.
First: calculating residual error rim, rimCalculation formula it is as follows:
rim=yi-Fm-1(xi) wherein, i=1 ..., n
Secondly, sampling gained training set with putting back to using the fuzzy learning machine algorithm that transfinitesTraining is fuzzy
Learning machine network transfinite as fm(x)。
Again, the fuzzy learning machine model F that transfinites of iteration is updatedm(x), specific formula is as follows:
Fm(x)=Fm-1(x)+fm(x)
Finally, calculating stop；If stop≤S, deconditioning exports Fm(x), otherwise continue, until m=M.Stop's
Calculation formula is as follows:
Backlight compensation method provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration, by using iteration mould
It pastes the learning machine that transfinites to return two backlight degree indexs of backlight image to be compensated, obtains the final of backlight image to be compensated
Backlight degree compensates backlight image to be compensated based on final backlight degree, to improve the brightness of backlight image, improves
The arithmetic speed of compensation expands the application range of backlight compensation.
On the basis of the above various embodiments, further, in the first backlight of the backlight image to be compensated that will acquire
Degree and the second backlight degree are input to the fuzzy learning machine that transfinites of trained iteration, before exporting third backlight degree, further includes:
The color space of the backlight image to be compensated is converted into YIQ by RGB；
First backlight degree is obtained using the brightness cumulative frequency histogram in the channel Y for the brightness value in the channel Y；
Using Da-Jin algorithm, second backlight degree is obtained.
Specifically, in detection-phase, the first backlight degree and the second backlight degree of backlight image to be compensated, backlight degree are obtained
For indicating the backlight degree of backlight image, and the degree of light filling is needed, the value of backlight degree reflects more greatly backlight image
Backlight degree is bigger, needs the degree of light filling also bigger.First backlight degree and the second backlight degree be primary Calculation obtain to
Compensate the backlight degree of backlight image.
Since the Y channel data in the YIQ color space of image has directly shown the brightness of image.Therefore, according to wait mend
The Y channel data for repaying backlight image calculates backlight degree.
Firstly, the color space of backlight image to be compensated is converted to YIQ by RGB, calculation formula is as follows:
Secondly, being directed to the brightness value in the channel Y, first is calculated come swept brilliance cumulative frequency polygon using sliding window
Backlight degree.
Second backlight degree is obtained finally, calculating using Da-Jin algorithm.
Backlight compensation method provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration, by using iteration mould
It pastes the learning machine that transfinites to return two backlight degree indexs of backlight image to be compensated, obtains the final of backlight image to be compensated
Backlight degree compensates backlight image to be compensated based on final backlight degree, to improve the brightness of backlight image, improves
The arithmetic speed of compensation expands the application range of backlight compensation.
It is further, described to be based on the third backlight degree to the back to be compensated on the basis of the above various embodiments
Light image compensates, and specifically includes:
According to the third backlight degree, default adaptive equalization curve break is determined；
The brightness value in the channel Y is compensated according to the default adaptive equalization curve；
The brightness value in the compensated channel Y is merged with the data in the channel I and the channel Q；
Color space is converted into RGB by YIQ again, obtains compensated backlight image.
Specifically, compensate that the specific method is as follows to backlight image to be compensated based on third backlight degree:
Firstly, determining default adaptive equalization curve break according to third backlight degree.
Adaptive equalization curve is preset further according to this to compensate the brightness value in the channel Y.
Then, the brightness value in the compensated channel Y is merged with the data in the channel I and the channel Q.
Finally, color space is converted to RGB by YIQ again, compensated backlight image is obtained.
Backlight compensation method provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration, by using iteration mould
It pastes the learning machine that transfinites to return two backlight degree indexs of backlight image to be compensated, obtains the final of backlight image to be compensated
Backlight degree compensates backlight image to be compensated based on final backlight degree, to improve the brightness of backlight image, improves
The arithmetic speed of compensation expands the application range of backlight compensation.
On the basis of the above various embodiments, further, the brightness cumulative frequency histogram using the channel Y is obtained
Take the calculation formula of first backlight degree are as follows:
Wherein, BhisFor the first backlight degree, Thi() is transfer function, SWmaxRefer to the brightness cumulative frequency histogram in the channel Y
In figure, when sliding window size and greater than min be less than max between pixel number account for total pixel number ratio it is equal when, max and
The maximum value at the interval between min, wherein max is the upper bound of sliding window, and min is the lower bound of sliding window.
Specifically, using the brightness cumulative frequency histogram in the channel Y, the first backlight degree is obtained.The meter of first backlight degree
Calculate formula are as follows:
Wherein, BhistFor the first backlight degree, Thist() is transfer function, SWmaxRefer to that the brightness cumulative frequency in the channel Y is straight
In square figure, when sliding window size and greater than min be less than max between pixel number account for total pixel number ratio it is equal when, max and
The maximum value at the interval between min, wherein max is the upper bound of sliding window, and min is the lower bound of sliding window.
Backlight compensation method provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration, by using iteration mould
It pastes the learning machine that transfinites to return two backlight degree indexs of backlight image to be compensated, obtains the final of backlight image to be compensated
Backlight degree compensates backlight image to be compensated based on final backlight degree, to improve the brightness of backlight image, improves
The arithmetic speed of compensation expands the application range of backlight compensation.
It is further, described to utilize Da-Jin algorithm on the basis of the above various embodiments, obtain the meter of second backlight degree
Calculate formula are as follows:
P (i)=ni/n
Wherein, B_Ostu is the second backlight degree, TOstu() is transfer function, C1For the minimum value of the brightness value in the channel Y,
C2For the maximum value of the brightness value in the channel Y, p (i) is the frequency that the pixel that brightness value is i occurs, niBe brightness value i pixel it is total
Number, n is sum of all pixels.
Specifically, using Da-Jin algorithm, the second backlight degree is obtained.The calculation formula of second backlight degree are as follows:
P (i)=ni/n
Wherein, B_Ostu is the second backlight degree, TOstu() is transfer function, C1And C2It is denoted as two of sample to be compensated
Cluster centre, C1For the minimum value of the brightness value in the channel Y, C2For the maximum value of the brightness value in the channel Y, p (i) is that brightness value is i
The frequency that pixel occurs, niIt is the sum of all pixels of brightness value i, n is sum of all pixels.
Backlight compensation method provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration, by using iteration mould
It pastes the learning machine that transfinites to return two backlight degree indexs of backlight image to be compensated, obtains the final of backlight image to be compensated
Backlight degree compensates backlight image to be compensated based on final backlight degree, to improve the brightness of backlight image, improves
The arithmetic speed of compensation expands the application range of backlight compensation.
On the basis of the above various embodiments, further, the formula of the default adaptive equalization curve are as follows:
Wherein, f (x) is default adaptive equalization curve, the coordinate for presetting adaptive equalization curve break be (a,
B), BdegreeFor third backlight degree, C1For the minimum value of the brightness value in the channel Y, C2For the maximum value of the brightness value in the channel Y.
Specifically, the brightness value in the channel Y is compensated according to default adaptive equalization curve.Default adaptive equalization
The formula of curve are as follows:
Wherein, f (x) is default adaptive equalization curve, the coordinate for presetting adaptive equalization curve break be (a,
B), BdegreeFor third backlight degree, C1For the minimum value of the brightness value in the channel Y, C2For the maximum value of the brightness value in the channel Y.
Backlight compensation method provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration, by using iteration mould
It pastes the learning machine that transfinites to return two backlight degree indexs of backlight image to be compensated, obtains the final of backlight image to be compensated
Backlight degree compensates backlight image to be compensated based on final backlight degree, to improve the brightness of backlight image, improves
The arithmetic speed of compensation expands the application range of backlight compensation.
Fig. 2 is the backlight compensation schematic device provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration, such as
Shown in Fig. 2, the embodiment of the present invention provides a kind of backlight compensation device based on the fuzzy learning machine that transfinites of iteration, above-mentioned for executing
Method described in any embodiment specifically includes computing module 201 and compensating module 202, in which:
The first backlight degree and the second backlight degree for the backlight image to be compensated that computing module 201 is used to will acquire, input
To the fuzzy learning machine that transfinites of trained iteration, third backlight degree is exported；Compensating module 202 is used to be based on the third backlight degree
The backlight image to be compensated is compensated.
Specifically, the compensation of the backlight compensation device provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration
Process includes detection-phase and compensated stage two parts.
Firstly, reading original backlight image to be compensated.
In detection-phase, the first backlight degree and the second backlight degree of backlight image to be compensated are obtained by computing module 201,
Backlight degree is used to indicate the backlight degree of backlight image, and needs the degree of light filling, and the value of backlight degree reflects more greatly backlight
The backlight degree of image is bigger, needs the degree of light filling also bigger.First backlight degree and the second backlight degree are that primary Calculation obtains
The backlight degree of the backlight image to be compensated arrived.
Then, the first backlight degree and the second backlight degree are input to the fuzzy learning machine that transfinites of trained iteration, output the
Three backlight degrees.That is, by using the fuzzy learning machine that transfinites of iteration to the first backlight degree and the second backlight of backlight image to be compensated
It is returned, obtains third backlight degree, which is the backlight degree of final, accurate backlight image to be compensated.Repeatedly
The generation fuzzy learning machine that transfinites is to be obtained on the basis of obscuring transfinites learning machine by stopping criterion for iteration when increasing training
, by the improvement to the fuzzy learning machine that transfinites, improve the robustness of the fuzzy learning machine that transfinites.
Backlight image to be compensated is compensated finally, being based on third backlight degree by compensating module 202.To improve
The local contrast of backlight image remains the details of image, and image is made to seem more natural, and in various ambient lighting conditions and
Image can easily and effectively be enhanced under angle, and improve compensation speed.
The embodiment of the present invention provides a kind of backlight compensation device based on the fuzzy learning machine that transfinites of iteration, above-mentioned for executing
Method described in any embodiment, the device provided through this embodiment execute above-mentioned a certain method as described in the examples
Specific steps are identical as above-mentioned corresponding embodiment, and details are not described herein again.
Backlight compensation device provided in an embodiment of the present invention based on the fuzzy learning machine that transfinites of iteration, by using iteration mould
It pastes the learning machine that transfinites to return two backlight degree indexs of backlight image to be compensated, obtains the final of backlight image to be compensated
Backlight degree compensates backlight image to be compensated based on final backlight degree, to improve the brightness of backlight image, improves
The arithmetic speed of compensation expands the application range of backlight compensation.
Fig. 3 is the structural schematic diagram of electronic equipment provided in an embodiment of the present invention, as shown in figure 3, the equipment includes: place
Manage device 301, memory 302 and bus 303；
Wherein, processor 301 and memory 302 complete mutual communication by the bus 303；
Processor 301 is used to call the program instruction in memory 302, to execute provided by above-mentioned each method embodiment
Method, for example,
The first backlight degree and the second backlight degree for the backlight image to be compensated that will acquire, are input to trained iteration mould
The learning machine that transfinites is pasted, third backlight degree is exported；
The backlight image to be compensated is compensated based on the third backlight degree.
The embodiment of the present invention provides a kind of computer program product, and the computer program product is non-transient including being stored in
Computer program on computer readable storage medium, the computer program include program instruction, when described program instructs quilt
When computer executes, computer is able to carry out method provided by above-mentioned each method embodiment, for example,
The first backlight degree and the second backlight degree for the backlight image to be compensated that will acquire, are input to trained iteration mould
The learning machine that transfinites is pasted, third backlight degree is exported；
The backlight image to be compensated is compensated based on the third backlight degree.
The embodiment of the present invention provides a kind of non-transient computer readable storage medium, the non-transient computer readable storage
Medium storing computer instruction, the computer instruction make the computer execute side provided by above-mentioned each method embodiment
Method, for example,
The first backlight degree and the second backlight degree for the backlight image to be compensated that will acquire, are input to trained iteration mould
The learning machine that transfinites is pasted, third backlight degree is exported；
The backlight image to be compensated is compensated based on the third backlight degree.
Those of ordinary skill in the art will appreciate that: realize that all or part of the steps of above method embodiment can pass through
The relevant hardware of program instruction is completed, and program above-mentioned can be stored in a computer readable storage medium, the program
When being executed, step including the steps of the foregoing method embodiments is executed；And storage medium above-mentioned includes: ROM, RAM, magnetic disk or light
The various media that can store program code such as disk.
The embodiments such as device and equipment described above are only schematical, wherein described be used as separate part description
Unit may or may not be physically separated, component shown as a unit may or may not be
Physical unit, it can it is in one place, or may be distributed over multiple network units.It can be according to the actual needs
Some or all of the modules therein is selected to achieve the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying
In the case where creative labor, it can understand and implement.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
It realizes by means of software and necessary general hardware platform, naturally it is also possible to pass through hardware.Based on this understanding, on
Stating technical solution, substantially the part that contributes to existing technology can be embodied in the form of software products in other words, should
Computer software product may be stored in a computer readable storage medium, such as ROM/RAM, magnetic disk, CD, including several fingers
It enables and using so that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation
Method described in certain parts of example or embodiment.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: it still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features；
And these are modified or replaceed, technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (10)
1. a kind of backlight compensation method based on the fuzzy learning machine that transfinites of iteration characterized by comprising
It is fuzzy super to be input to trained iteration for the first backlight degree and the second backlight degree for the backlight image to be compensated that will acquire
Learning machine is limited, third backlight degree is exported；
The backlight image to be compensated is compensated based on the third backlight degree.
2. the method according to claim 1, wherein the training method of the fuzzy learning machine that transfinites of the iteration is such as
Under:
Obtain the preset stopping condition and default the number of iterations of the fuzzy learning machine that transfinites of training；
Several training samples that will have been marked in advance are input to the fuzzy learning machine that transfinites and are iterated training, work as the number of iterations
When meeting the preset stopping condition when reaching the default the number of iterations or certain iteration, then training terminates, and is trained
The fuzzy learning machine that transfinites of good iteration.
3. according to the method described in claim 2, it is characterized in that, the backlight image to be compensated that will acquire the first backlight
Degree and the second backlight degree are input to the fuzzy learning machine that transfinites of trained iteration, before exporting third backlight degree, further includes:
The color space of the backlight image to be compensated is converted into YIQ by RGB；
First backlight degree is obtained using the brightness cumulative frequency histogram in the channel Y for the brightness value in the channel Y；
Using Da-Jin algorithm, second backlight degree is obtained.
4. according to the method described in claim 3, it is characterized in that, described be based on the third backlight degree to the back to be compensated
Light image compensates, and specifically includes:
According to the third backlight degree, default adaptive equalization curve break is determined；
The brightness value in the channel Y is compensated according to the default adaptive equalization curve；
The brightness value in the compensated channel Y is merged with the data in the channel I and the channel Q；
Color space is converted into RGB by YIQ again, obtains compensated backlight image.
5. according to the method described in claim 3, it is characterized in that, the brightness cumulative frequency histogram using the channel Y, is obtained
Take the calculation formula of first backlight degree are as follows:
Wherein, BhistFor the first backlight degree, Thist() is transfer function, SWmaxRefer to the brightness cumulative frequency histogram in the channel Y
In, when sliding window size and greater than min be less than the pixel number between max account for total pixel number ratio it is equal when, max and min
Between interval maximum value, wherein max be sliding window the upper bound, min be sliding window lower bound.
6. according to the method described in claim 3, obtaining second backlight degree it is characterized in that, described utilize Da-Jin algorithm
Calculation formula are as follows:
P (i)=ni/n
Wherein, B_Ostu is the second backlight degree, TOstu() is transfer function, C1For the minimum value of the brightness value in the channel Y, C2For Y
The maximum value of the brightness value in channel, p (i) are the frequency that the pixel that brightness value is i occurs, niIt is the sum of all pixels of brightness value i, n
It is sum of all pixels.
7. according to the method described in claim 4, it is characterized in that, the formula of the default adaptive equalization curve are as follows:
Wherein, f (x) is default adaptive equalization curve, and the coordinate for presetting adaptive equalization curve break is (a, b),
BdegreeFor third backlight degree, C1For the minimum value of the brightness value in the channel Y, C2For the maximum value of the brightness value in the channel Y.
8. a kind of backlight compensation device based on the fuzzy learning machine that transfinites of iteration characterized by comprising
Computing module, the first backlight degree and the second backlight degree of the backlight image to be compensated for will acquire, is input to training
The fuzzy learning machine that transfinites of good iteration, exports third backlight degree；
Compensating module, for being compensated based on the third backlight degree to the backlight image to be compensated.
9. a kind of electronic equipment characterized by comprising
Memory and processor, the processor and the memory complete mutual communication by bus；The memory
It is stored with the program instruction that can be executed by the processor, the processor calls described program instruction to be able to carry out right such as and wants
Seek 1 to 7 any method.
10. a kind of non-transient computer readable storage medium, is stored thereon with computer program, which is characterized in that when the meter
When calculation machine program is executed by processor, the method as described in claim 1 to 7 is any is realized.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811337780.3A CN109584172A (en) | 2018-11-12 | 2018-11-12 | Backlight compensation method and device based on the fuzzy learning machine that transfinites of iteration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811337780.3A CN109584172A (en) | 2018-11-12 | 2018-11-12 | Backlight compensation method and device based on the fuzzy learning machine that transfinites of iteration |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109584172A true CN109584172A (en) | 2019-04-05 |
Family
ID=65922164
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811337780.3A Withdrawn CN109584172A (en) | 2018-11-12 | 2018-11-12 | Backlight compensation method and device based on the fuzzy learning machine that transfinites of iteration |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109584172A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113674705A (en) * | 2021-08-27 | 2021-11-19 | 天津大学 | Backlight extraction method based on radial basis function neural network proxy model assisted particle swarm algorithm |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106161931A (en) * | 2016-06-28 | 2016-11-23 | 广东欧珀移动通信有限公司 | The method and device of image preview |
| CN106845458A (en) * | 2017-03-05 | 2017-06-13 | 北京工业大学 | A kind of rapid transit label detection method of the learning machine that transfinited based on core |
| CN106960176A (en) * | 2017-02-22 | 2017-07-18 | 华侨大学 | A kind of pedestrian's gender identification method based on transfinite learning machine and color characteristic fusion |
| CN107135353A (en) * | 2017-04-27 | 2017-09-05 | 努比亚技术有限公司 | A kind of exposure regulating method, mobile terminal and computer-readable recording medium |
-
2018
- 2018-11-12 CN CN201811337780.3A patent/CN109584172A/en not_active Withdrawn
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106161931A (en) * | 2016-06-28 | 2016-11-23 | 广东欧珀移动通信有限公司 | The method and device of image preview |
| CN106960176A (en) * | 2017-02-22 | 2017-07-18 | 华侨大学 | A kind of pedestrian's gender identification method based on transfinite learning machine and color characteristic fusion |
| CN106845458A (en) * | 2017-03-05 | 2017-06-13 | 北京工业大学 | A kind of rapid transit label detection method of the learning machine that transfinited based on core |
| CN107135353A (en) * | 2017-04-27 | 2017-09-05 | 努比亚技术有限公司 | A kind of exposure regulating method, mobile terminal and computer-readable recording medium |
Non-Patent Citations (1)
| Title |
|---|
| 王聪聪: "复杂光照条件下交通标识检测算法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113674705A (en) * | 2021-08-27 | 2021-11-19 | 天津大学 | Backlight extraction method based on radial basis function neural network proxy model assisted particle swarm algorithm |
| CN113674705B (en) * | 2021-08-27 | 2023-11-07 | 天津大学 | Backlight extraction method based on radial basis function neural network agent model auxiliary particle swarm algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lv et al. | Attention guided low-light image enhancement with a large scale low-light simulation dataset | |
| Zhuang et al. | Underwater image enhancement with hyper-laplacian reflectance priors | |
| Gharbi et al. | Deep bilateral learning for real-time image enhancement | |
| US9275445B2 (en) | High dynamic range and tone mapping imaging techniques | |
| CN102905058B (en) | Produce the apparatus and method for eliminating the fuzzy high dynamic range images of ghost image | |
| Al‐Ameen | Nighttime image enhancement using a new illumination boost algorithm | |
| US20200234414A1 (en) | Systems and methods for transforming raw sensor data captured in low-light conditions to well-exposed images using neural network architectures | |
| EP3542347B1 (en) | Fast fourier color constancy | |
| US9292911B2 (en) | Automatic image adjustment parameter correction | |
| US9070044B2 (en) | Image adjustment | |
| CN112200736B (en) | Image processing method based on reinforcement learning and model training method and device | |
| US8787659B2 (en) | Automatic adaptation to image processing pipeline | |
| CN108668093A (en) | The generation method and device of HDR image | |
| Wang et al. | Variational single nighttime image haze removal with a gray haze-line prior | |
| WO2023160312A1 (en) | Person re-identification method and apparatus based on self-supervised learning, and device and storage medium | |
| CN110570435A (en) | method and device for carrying out damage segmentation on vehicle damage image | |
| Smelyakov et al. | Comparative efficiency analysis of gradational correction models of highly lighted image | |
| CN109325918A (en) | Image processing method and device and computer storage medium | |
| US11138693B2 (en) | Attention-driven image manipulation | |
| Yang et al. | Improved single image dehazing methods for resource-constrained platforms | |
| Hsieh et al. | Variational contrast-saturation enhancement model for effective single image dehazing | |
| CN106611406B (en) | Method for correcting image and image correcting device | |
| Yan et al. | A natural-based fusion strategy for underwater image enhancement | |
| CN109584172A (en) | Backlight compensation method and device based on the fuzzy learning machine that transfinites of iteration | |
| Zhang et al. | An adaptive color correction method for underwater single image haze removal |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190405 |