CN106780400A - A kind of image processing method and device - Google Patents
A kind of image processing method and device Download PDFInfo
- Publication number
- CN106780400A CN106780400A CN201710021180.5A CN201710021180A CN106780400A CN 106780400 A CN106780400 A CN 106780400A CN 201710021180 A CN201710021180 A CN 201710021180A CN 106780400 A CN106780400 A CN 106780400A
- Authority
- CN
- China
- Prior art keywords
- frequency
- pixel
- image
- exploded view
- width
- 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 13
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 77
- 238000000034 method Methods 0.000 claims abstract description 50
- 230000009466 transformation Effects 0.000 claims abstract description 11
- 238000001914 filtration Methods 0.000 claims description 87
- 238000003384 imaging method Methods 0.000 claims description 21
- 230000008569 process Effects 0.000 claims description 20
- 238000010276 construction Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 13
- 238000004590 computer program Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000012986 modification Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 241000208340 Araliaceae Species 0.000 description 2
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 2
- 235000003140 Panax quinquefolius Nutrition 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 235000008434 ginseng Nutrition 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 241001270131 Agaricus moelleri Species 0.000 description 1
- 238000004519 manufacturing process 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
- G06T5/70—Denoising; Smoothing
-
- 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
- G06T2207/20024—Filtering details
-
- 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
- G06T2207/20064—Wavelet transform [DWT]
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of image processing method and device, and the method includes：Using wavelet transformation, the original brightness image to original image carries out N rank decomposition；For every width exploded view picture, according to the width exploded view as corresponding frequency adjustment threshold value is adjusted to the corresponding frequency of each pixel in the width exploded view picture, wherein, the frequency more than frequency adjustment threshold value is turned down, and the frequency less than frequency adjustment threshold value is heightened；According to the every width exploded view picture after frequency adjustment, luminance picture after the corresponding adjustment of the original image is reconstructed；According to the original image, original brightness image and luminance picture after adjustment, image after the corresponding treatment of the original image is determined.The present invention reproduced image details and can strengthen the details in the region that details in image is not protruded, additionally it is possible to lift original image dark place image detail and reproduce the details of highlight regions image.
Description
Technical field
The present invention relates to digital image processing techniques field, more particularly to a kind of image processing method and device.
Background technology
In image display arts, usual display can show pixel of the brightness value of R, G, B component between [0,255]
Value, and 256 different gray scales obviously can not completely reflect the brightness case of light in nature.Such as, the sun is bright
Degree is probably one thousands of times of incandescent lamp, is a hundreds of thousands for the brightness of the desktop illuminated by incandescent lamp times, and this is remote
Far beyond the display brightness range to be shown, the image for now being shown on display it is possible that bright areas expose
Light is excessive, and dark goes under-exposure, leads to not the problem of resolution image details.Therefore, how in limited brightness range
The so broad brightness range of interior display, exactly image processing techniques problem to be solved.
At present, Tone Mapping algorithms are commonly used, image detail is lifted, so as to shows broadness in limited brightness range
Brightness range image, and then lifted picture quality.The mainly method drawn game including Global treatment of Tone Mapping algorithms
The method of portion's treatment.The method of Global treatment is main by using each pixel in same nonlinear curve mapping graph picture,
So as to reach the purpose of lifting image detail；The method of Local treatment is based primarily upon Retinex theories, the theoretical basic thought
By the illumination value around the absolute illumination value and the scene that consider the scene that image capture device is collected, image is entered
Row treatment of details.
But, the method based on Global treatment cannot accomplish to reproduce highlight bar while lifting dark place image detail is ensured
The details of area image, therefore, the picture contrast after being processed using the method is low relatively low, and image detail is also not clear enough.
And because the change of Retinex theory hypothesis image irradiation is gentle, therefore use the method for Local treatment to hold when image is processed
Easily " halation phenomenon " is being produced in the strong place of comparison of light and shade.
The content of the invention
The present invention provides a kind of image processing method and device, for solving to carry out treatment of details to image in the prior art
The low relatively low, image detail of picture contrast that is brought of mode it is also not clear enough and easily produce " halation phenomenon "
Problem.
A kind of image processing method, including：
Using wavelet transformation, the original brightness image to original image carries out N rank decomposition, and N is the integer not less than 1；
For every width exploded view picture, according to the width exploded view as corresponding frequency adjusts threshold value to every in the width exploded view picture
The corresponding frequency of individual pixel is adjusted, wherein, the frequency more than frequency adjustment threshold value is turned down, less than frequency adjustment threshold value
Frequency is heightened；
According to the every width exploded view picture after frequency adjustment, luminance picture after the corresponding adjustment of the original image is reconstructed；
According to the original image, original brightness image and luminance picture after adjustment, the original image correspondence is determined
Treatment after image.
Alternatively, in methods described, using wavelet transformation, the luminance picture to original image carries out N rank decomposition, specific bag
Include：
Using Stationary Wavelet Transform, the original brightness image to original image carries out N rank decomposition, obtains 3N width high-frequency decompositions
Image and 1 width low frequency exploded view picture, wherein, being decomposed per rank and obtain 3 width high-frequency decomposition images, last single order is decomposed and also obtains 1
Width low frequency exploded view picture.
Alternatively, in methods described, according to the width exploded view as corresponding frequency adjusts threshold value in the width exploded view picture
The corresponding frequency of each pixel is adjusted, and specifically includes：
According to the corresponding frequency of each pixel in the width exploded view picture, and the width exploded view is adjusted as corresponding frequency
Threshold value, determines the corresponding gain of each pixel in the width exploded view picture；
According to the corresponding gain of each pixel in the width exploded view picture, to corresponding pixel points correspondence in the width exploded view picture
Frequency be adjusted.
Alternatively, in methods described, determine the corresponding gain of each pixel in the width exploded view picture, specifically include：
For each pixel in the width exploded view picture, gaussian filtering process is carried out to the corresponding frequency of the pixel, obtained
Frequency after to the corresponding filtering of the pixel；
Determine each corresponding picture of position in other each width exploded view pictures with the pixel in the width exploded view picture
Frequency after the corresponding filtering of vegetarian refreshments；
Calculate after the corresponding filtering of the pixel in frequency and other each width exploded view pictures with the pixel in the width point
The average value of frequency after the corresponding corresponding filtering of each pixel in position in solution image, as the corresponding filter of the pixel
Average frequency value after ripple；
According to the width exploded view as average frequency value after corresponding frequency adjustment threshold value and the corresponding filtering of the pixel,
Calculate the corresponding gain of the pixel.
The embodiment of the present invention, after the corresponding frequency of pixel is carried out into gaussian filtering process, can in removable original image
The halation phenomenon that can occur, additionally it is possible to lift the details of image so that image detail is more prominent.
Alternatively, in methods described, gaussian filtering process is carried out to the corresponding frequency of the pixel, obtains the pixel pair
Frequency after the filtering answered, specifically includes：
If the width exploded view picture is any rank decomposes any high-frequency decomposition image for obtaining, determine that any rank is decomposed
The other two width high-frequency decomposition image for arriving；
Determine the position phase with the pixel in any high-frequency decomposition image in the other two width high frequency imaging
The corresponding frequency of corresponding two pixels；
The absolute value of absolute value and the corresponding frequency of described two pixels according to the corresponding frequency of the pixel, meter
Calculate the corresponding average frequency value of the pixel；
Gaussian filtering process is carried out to the corresponding average frequency value of the pixel, frequency after the corresponding filtering of the pixel is obtained
Rate.
The embodiment of the present invention, after the corresponding frequency of pixel is carried out into gaussian filtering process, can in removable original image
The halation phenomenon that can occur, additionally it is possible to lift the details of image so that image detail is more prominent.
Alternatively, in methods described, the corresponding gain of the pixel is calculated using equation below, is specifically included：
Wherein, i represents decomposition exponent number, and j represents that i ranks decompose any width exploded view picture for obtaining, Gij(x, y) represents i ranks point
The exploded view that solution is obtained is as the corresponding gain of pixel in (x, y) position, A in jag(x, y) represents exploded view as being in j
Average frequency value, m after the corresponding filtering of pixel of (x, y) positioniRepresent that i ranks decompose corresponding gain control parameter and 0≤mi
≤ 1, γ are preset constant and 0≤γ≤1, and δ is default constant and 0≤δ≤1, and A' is exploded view as each pixel pair in j
The average value of the frequency answered, ε is for suppressing the exaggerated noise level relevant parameter of noise and 0≤ε≤1, Ak(x, y) is represented
Carry out in the kth width image in the 3N+1 width images obtained after N rank decomposition after the corresponding filtering of pixel in (x, y) position
Frequency.
Alternatively, in methods described, determine image after the corresponding treatment of details of the original image, specifically include：
For each the original image vegetarian refreshments in the original image, the position of the original image vegetarian refreshments in original image is determined
Put；
According to the original of pixel corresponding with the position of the original image vegetarian refreshments in original image in original brightness image
The tune of pixel corresponding with the position of the original image vegetarian refreshments in original image in luminance picture after beginning brightness, and adjustment
Whole rear brightness, is adjusted to the original pixels point；
Built according to the pixel after each adjustment in original image and obtain image after the corresponding treatment of original image.
The embodiment of the present invention, each original pixels point for original image carries out treatment of details, so as to reach to original
Image carries out the purpose for the treatment of of details.
Alternatively, in methods described, the original pixels point is adjusted using equation below：
Wherein, Cin(x, y) represents the raw image data of the original image vegetarian refreshments in (x, y) position in original image, Cout
(x, y) represents the view data after being adjusted to the original pixels point in (x, y) position in original image, Iout(x, y) table
Show brightness, I after the adjustment of the pixel in (x, y) position in luminance picture after adjustingin(x, y) is represented in original brightness image
The original brightness of the pixel in (x, y) position.
The present invention also provides a kind of image processing apparatus, including：
Decomposing module, for utilizing wavelet transformation, the original brightness image to original image carries out N rank decomposition, and N is not small
In 1 integer；
Frequency regulation block, for for every width exploded view picture, according to the width exploded view as corresponding frequency adjusts threshold value
The corresponding frequency of each pixel in the width exploded view picture is adjusted, wherein, the frequency more than frequency adjustment threshold value is turned down,
Frequency less than frequency adjustment threshold value is heightened；
Reconstructed module, for being adjusted according to frequency after every width exploded view picture, reconstruct the corresponding tune of the original image
Whole rear luminance picture；
Determining module, for according to the original image, original brightness image and luminance picture after adjustment, it is determined that described
Image after the corresponding treatment of original image.
Alternatively, in described device, the decomposing module, specifically for：
Using Stationary Wavelet Transform, the original brightness image to original image carries out N rank decomposition, obtains 3N width high-frequency decompositions
Image and 1 width low frequency exploded view picture, wherein, being decomposed per rank and obtain 3 width high-frequency decomposition images, last single order is decomposed and also obtains 1
Width low frequency exploded view picture.
Alternatively, in described device, the frequency regulation block is specifically included：
Gain determination unit, for being decomposed according to the corresponding frequency of each pixel in the width exploded view picture, and the width
The corresponding frequency adjustment threshold value of image, determines the corresponding gain of each pixel in the width exploded view picture；
Frequency adjustment unit, for according to the corresponding gain of each pixel in the width exploded view picture, to the width exploded view
The corresponding frequency of corresponding pixel points is adjusted as in.
Alternatively, in described device, the gain determination unit, specifically for：
For each pixel in the width exploded view picture, gaussian filtering process is carried out to the corresponding frequency of the pixel, obtained
Frequency after to the corresponding filtering of the pixel；
Determine each corresponding picture of position in other each width exploded view pictures with the pixel in the width exploded view picture
Frequency after the corresponding filtering of vegetarian refreshments；
Calculate after the corresponding filtering of the pixel in frequency and other each width exploded view pictures with the pixel in the width point
The average value of frequency after the corresponding corresponding filtering of each pixel in position in solution image, as the corresponding filter of the pixel
Average frequency value after ripple；
According to the width exploded view as average frequency value after corresponding frequency adjustment threshold value and the corresponding filtering of the pixel,
Calculate the corresponding gain of the pixel.
Alternatively, in described device, the gain determination unit, specifically for：
If the width exploded view picture is any rank decomposes any high-frequency decomposition image for obtaining, determine that any rank is decomposed
The other two width high-frequency decomposition image for arriving；
Determine the position phase with the pixel in any high-frequency decomposition image in the other two width high frequency imaging
The corresponding frequency of corresponding two pixels；
The absolute value of absolute value and the corresponding frequency of described two pixels according to the corresponding frequency of the pixel, meter
Calculate the corresponding average frequency value of the pixel；
Gaussian filtering process is carried out to the corresponding average frequency value of the pixel, frequency after the corresponding filtering of the pixel is obtained
Rate.
Alternatively, in described device, the gain determination unit using equation below specifically for calculating the pixel pair
The gain answered：
Wherein, i represents decomposition exponent number, and j represents that i ranks decompose any width exploded view picture for obtaining, Gij(x, y) represents i ranks point
The exploded view that solution is obtained is as the corresponding gain of pixel in (x, y) position, A in jag(x, y) represents exploded view as being in j
Average frequency value, m after the corresponding filtering of pixel of (x, y) positioniRepresent that i ranks decompose corresponding gain control parameter and 0≤mi
≤ 1, γ are preset constant and 0≤γ≤1, and δ is default constant and 0≤δ≤1, and A' is exploded view as each pixel pair in j
The average value of the frequency answered, ε is for suppressing the exaggerated noise level relevant parameter of noise and 0≤ε≤1, Ak(x, y) is represented
Carry out in the kth width image in the 3N+1 width images obtained after N rank decomposition after the corresponding filtering of pixel in (x, y) position
Frequency.
Alternatively, in described device, the determining module is specifically included：
Position determination unit, for for each the original image vegetarian refreshments in the original image, determining the original image vegetarian refreshments
Position in original image；
Brightness adjustment unit, for according to the position phase in original brightness image with the original image vegetarian refreshments in original image
Position phase after the original brightness of corresponding pixel, and adjustment in luminance picture with the original image vegetarian refreshments in original image
Brightness after the adjustment of corresponding pixel, is adjusted to the original pixels point；
Construction unit, the corresponding place of original image is obtained for being built according to the pixel after each adjustment in original image
Image after reason.
Alternatively, in described device, the brightness adjustment unit is specifically for using equation below to the original image vegetarian refreshments
It is adjusted：
Wherein, Cin(x, y) represents the raw image data of the original image vegetarian refreshments in (x, y) position in original image, Cout
(x, y) represents the view data after being adjusted to the original pixels point in (x, y) position in original image, Iout(x, y) table
Show brightness, I after the adjustment of the pixel in (x, y) position in luminance picture after adjustingin(x, y) is represented in original brightness image
The original brightness of the pixel in (x, y) position.
Using image processing method provided in an embodiment of the present invention and device, have the advantages that：By will be original
The luminance picture of image is decomposed, and obtains the exploded view picture comprising frequency, and the frequency to each exploded view picture is adjusted, and
Image Reconstruction is decomposed after being adjusted according to frequency and obtains image after the corresponding adjustment of original image, so as to reach reproduced image details,
And strengthen the purpose of details in the region that details in image is not protruded；According to original brightness image, whole rear luminance picture to original
Beginning image is processed, and so as to obtain image after the corresponding treatment of original image, to improve original image dark for image after the treatment
Place's image detail, and reproduced the details of highlight regions image.
Brief description of the drawings
Fig. 1 is image processing method schematic flow sheet provided in an embodiment of the present invention；
Fig. 2 is the method that the corresponding frequency of each pixel is adjusted in the picture to exploded view provided in an embodiment of the present invention
Schematic flow sheet；
Fig. 3 is that the method flow of the corresponding gain of each pixel in determination exploded view picture provided in an embodiment of the present invention shows
It is intended to；
Fig. 4 is the method flow schematic diagram for obtaining frequency after the corresponding filtering of pixel provided in an embodiment of the present invention；
Fig. 5 a are width high-frequency decomposition image schematic diagram provided in an embodiment of the present invention；
Fig. 5 b are another width high-frequency decomposition image schematic diagram provided in an embodiment of the present invention；
Fig. 5 c are another width high-frequency decomposition image schematic diagram provided in an embodiment of the present invention；
Fig. 6 is the method flow signal of image after the corresponding treatment of details of determination original image provided in an embodiment of the present invention
Figure；
Fig. 7 is image processing apparatus structural representation provided in an embodiment of the present invention.
Specific embodiment
The image processing method and device for providing the present invention with reference to the accompanying drawings and examples are illustrated in greater detail.
The embodiment of the present invention provides a kind of image processing method, as shown in figure 1, the method includes：
Step 101, using wavelet transformation, the original brightness image to original image carries out N rank decomposition, and N is not less than 1
Integer.
Specifically, N rank decomposition can be carried out to the original brightness image of original image using wavelet transformation, such as can be using just
Hand over wavelet transformation or Stationary Wavelet Transform, it is preferable that the original brightness image of original image is entered using Stationary Wavelet Transform
Row N ranks are decomposed.
Wherein, after carrying out N rank decomposition to the original brightness image of original image using Stationary Wavelet Transform, resulting divides
The size of image and the size of original image are identical after solution, also, utilization Stationary Wavelet Transform is to the original brightness of original image
After image carries out N rank decomposition, 3N width high-frequency decomposition images and 1 width low frequency exploded view picture are obtained, wherein, decomposed per rank and obtain 3
Width high-frequency decomposition image, last single order is decomposed and also obtains 1 width low frequency exploded view picture, and high frequency, low frequency are included in high-frequency decomposition image
Low frequency is included in exploded view picture.
It is specific that the process of N rank decomposition is carried out to original image reference can be made to prior art using Stationary Wavelet Transform, here
It is not detailed.Wherein, the frequency comprising image in the exploded view picture for obtaining is decomposed per rank.
It should be noted that the corresponding view data of original image can be Bayer data, yuv data or RGB numbers
According to.Monochrome information in each pixel of extractable original picture, constitutes the luminance picture of original image, specifically obtains original
The mode of the luminance picture of image can be found in prior art, be not detailed here.
Step 102, for every width exploded view picture, decomposes as corresponding frequency adjusts threshold value according to the width exploded view to the width
The corresponding frequency of each pixel is adjusted in image, wherein, the frequency more than frequency adjustment threshold value is turned down, and is adjusted less than frequency
The frequency of whole threshold value is heightened.
Specifically, decomposing each width exploded view for obtaining per rank as corresponding frequency adjustment threshold value can be according to practical application scene
Setting, does not limit here, and the corresponding frequency adjustment threshold value of different decomposition image identical may be likely to difference.
Specifically, can reflect image detail for the high frequency of piece image, low frequency reacts image outline, in this step, will
The frequency that frequency adjustment threshold value is will be greater than in exploded view picture carries out turning down treatment, highlight regions in image can be dimmed into reproduction highlighted
The details of area image, the frequency that frequency adjustment threshold value will will be less than in exploded view picture carry out turning down treatment, can will be low in image
Bright area lightens and strengthens the details in the region that details in image is not protruded.
Step 103, according to the every width exploded view picture after frequency adjustment, reconstructs bright after the corresponding adjustment of the original image
Degree image.
Specifically, each width exploded view picture after being adjusted to frequency is reconstructed, obtain bright after the corresponding adjustment of original image
Degree image.
Step 104, according to the original image, original brightness image and luminance picture after adjustment, determines described original
Image after the corresponding treatment of image.
Specifically, calculate the ratio of luminance picture and original brightness image after adjustment, ratio is multiplied institute with original image
The result images for obtaining are used as image after the corresponding treatment of original image.
The embodiment of the present invention, is decomposed by by the luminance picture of original image, obtains the exploded view picture comprising frequency,
Frequency to each exploded view picture is adjusted, and obtains the corresponding tune of original image according to Image Reconstruction is decomposed after frequency adjustment
Whole rear image, so as to reach reproduced image details, and strengthens the purpose of details in the region that details in image is not protruded；According to
Original brightness image, whole rear luminance picture are processed original image, so as to obtain image after the corresponding treatment of original image,
Image improves original image dark place image detail after the treatment, and has reproduced the details of highlight regions image.
Preferably, the content for being provided according to Fig. 2, decomposes as corresponding frequency adjusts threshold value according to the width exploded view to the width
The corresponding frequency of each pixel is adjusted in image：
Step 201, according to the corresponding frequency of each pixel in the width exploded view picture, and the width exploded view is as corresponding
Frequency adjusts threshold value, determines the corresponding gain of each pixel in the width exploded view picture.
Step 202, according to the corresponding gain of each pixel in the width exploded view picture, to correspondence picture in the width exploded view picture
The corresponding frequency of vegetarian refreshments is adjusted.
Specifically, determine the corresponding gain of each pixel in a width exploded view picture, so as to obtain the width exploded view picture pair
The gain relationship answered, the gain relationship is the corresponding relation of the yield value corresponding to the pixel of diverse location in image, wherein,
The corresponding gain of different pixels point identical may be likely to difference in same width exploded view picture, and this is big by the corresponding frequency of pixel
It is small to be determined.
Wherein, the corresponding frequency of each pixel in the width exploded view picture is adjusted, is specifically included：For the width point
Each pixel in solution image, by the corresponding gain of pixel frequency multiplication corresponding with the pixel, obtains frequency adjustment
Pixel afterwards.
Preferably, the content for being provided according to Fig. 3, determines the corresponding gain of each pixel in the width exploded view picture：
Step 301, for each pixel in the width exploded view picture, gaussian filtering is carried out to the corresponding frequency of the pixel
Treatment, obtains frequency after the corresponding filtering of the pixel.
Step 302, determines that the position in other each width exploded view pictures with the pixel in the width exploded view picture is corresponding
The corresponding filtering of each pixel after frequency.
Specifically, other each width exploded view pictures are the exploded view picture outside the width exploded view picture, other each width exploded view pictures
In corresponding each pixel in position with the pixel in the width exploded view picture, refer specifically to, other each width exploded view pictures
In there is the pixel of same position with the pixel, the position of the pixel is specially the pixel in the width exploded view picture
Position, the explanation of specific same position can be found in the explanation for Fig. 5 a, Fig. 5 b and the corresponding examples of Fig. 5 c, do not do here
Describe in detail.
Step 303, calculate after the corresponding filtering of the pixel in frequency and other each width exploded view pictures with the pixel
The average value of frequency after the corresponding corresponding filtering of each pixel in position in the width exploded view picture, as the pixel
Average frequency value after corresponding filtering.
Using the embodiment of the present invention, average frequency value after the corresponding filtering of pixel of same position in each width exploded view picture
It is identical.
Step 304, according to the width exploded view as frequency after corresponding frequency adjustment threshold value and the corresponding filtering of the pixel
Rate average value, calculates the corresponding gain of the pixel.
In this preferred embodiment, for each pixel in every width exploded view picture, the corresponding frequency of the pixel is entered
Row gaussian filtering process, so as to calculate the corresponding gain of the pixel according to frequency after filtering.
The embodiment of the present invention, after the corresponding frequency of pixel is carried out into gaussian filtering process, can in removable original image
The halation phenomenon that can occur, additionally it is possible to lift the details of image so that image detail is more prominent.
Wherein, for every width exploded view picture, gaussian filtering is carried out to the corresponding frequency of each pixel in the width image
Treatment, obtains frequency after the corresponding filtering of each pixel in the width image.
Using the embodiment of the present invention, the 3 width high frequency imagings for obtaining are decomposed per rank may correspond to same gain relationship, may be right
Different gains relation is answered, last single order decomposes low-frequency image one gain relationship of correspondence for obtaining, or every width exploded view picture difference
One gain relationship of correspondence.
The embodiment that Fig. 3 is provided is only a kind of preferred embodiment, also can in the following way determine the width exploded view
The corresponding gain of each pixel as in：For each pixel in the width exploded view picture, the corresponding frequency of the pixel is entered
Row gaussian filtering process, obtains frequency after the corresponding filtering of the pixel；According to the width exploded view as corresponding frequency adjusts threshold
Frequency after value and the corresponding filtering of the pixel, calculates the corresponding gain of the pixel.Need not pixel be calculated corresponding
Average frequency value after filtering, directly according to frequency after the corresponding filtering of pixel and the width exploded view as corresponding frequency is adjusted
The corresponding gain of threshold calculations pixel.
Preferably, the content for being provided according to Fig. 4, gaussian filtering process is carried out to the corresponding frequency of the pixel, is somebody's turn to do
Frequency after the corresponding filtering of pixel：
Step 401, if the width exploded view picture is any rank decomposes any high-frequency decomposition image for obtaining, determines described any
Rank decomposes the other two width high-frequency decomposition image for obtaining.
Specifically, after N rank decomposition is carried out to the original brightness image of original image using Stationary Wavelet Transform, obtaining 3N
Width high-frequency decomposition image and 1 width low frequency exploded view picture, in this step in the specific implementation, predefine the currently processed width
Exploded view seems no for any rank decomposes any high-frequency decomposition image for obtaining, if it is, determine that any rank is decomposed obtaining
Other two width high-frequency decomposition image, if it is not, then determining that the currently processed width exploded view picture is obtained for last single order is decomposed
Low-frequency image, and for each pixel in the low-frequency image, the corresponding frequency of the pixel is carried out at gaussian filtering
Reason, and according to frequency after the corresponding filtering of the pixel, calculate the corresponding gain of the pixel.
Step 402, determines in the other two width high frequency imaging with the pixel in any high-frequency decomposition image
The corresponding corresponding frequency of two pixels in position.
Wherein, the position phase in the other two width high frequency imaging with the pixel in any high-frequency decomposition image
Corresponding two pixels, refer specifically to, the pixel with the pixel in the other two width high frequency imaging with same position,
The position of the pixel is specially position of the pixel in any high-frequency decomposition image.
Specifically, the pixel number included in every width exploded view picture is identical, and each width exploded view pixel as in
Arrangement mode is identical.In this step, position of the pixel in any high-frequency decomposition image is determined, and then for another
Every width high frequency imaging in outer two width high frequency imaging, determines the picture with same position with the pixel in the width high frequency imaging
The corresponding high-frequency information of vegetarian refreshments.For example, be 3 width high-frequency decomposition image schematic diagrames as shown in Fig. 5 a, Fig. 5 b and Fig. 5 c, its
In, Fig. 5 a are any high-frequency decomposition image, including 4 pixels a1, a2, a3, a4, Fig. 5 b and Fig. 5 c are respectively and this
One high-frequency decomposition image decomposes the other two width high frequency imaging for obtaining in same single order, respectively including 4 pixels, the picture in Fig. 5 b
Vegetarian refreshments is b1, and the pixel in b2, b3, b4, Fig. 5 c is c1, c2, c3, c4, then, with pixel a1 there is same position (to be in
Same position) other two width high frequency imaging in pixel be respectively b1 and c1, there is the another of same position with pixel a2
Pixel in outer two width high frequency imaging is respectively b2 and c2, with the other two width high frequency imaging that pixel a3 has same position
In pixel be respectively b3 and c3, with pixel a4 there is pixel in the other two width high frequency imaging of same position to distinguish
It is b4 and c4.
Specifically, Fig. 5 a, Fig. 5 b and Fig. 5 c are located at same two-dimensional coordinate system, and Fig. 5 a, Fig. 5 b and Fig. 5 c overlap, pair with should
For two-dimensional coordinate system, the pixel in (x, y) position includes three pixels, and these three pixels are respectively Fig. 5 a, figure
Pixel with same position in 5b and Fig. 5 c.
Step 403, absolute value and the corresponding frequency of described two pixels according to the corresponding frequency of the pixel
Absolute value, calculates the corresponding average frequency value of the pixel.
Specifically, in calculating the absolute value F1 of the corresponding frequency of the pixel, and other two width high frequency imaging, with the picture
The absolute value F2 and F3 of the corresponding corresponding frequency of two pixels in position of the vegetarian refreshments in any high-frequency decomposition image.
The computing formula of the corresponding average frequency value F of the pixel is：
Step 404, gaussian filtering process is carried out to the corresponding average frequency value of the pixel, obtains the pixel corresponding
Frequency after filtering.
The embodiment of the present invention, after the corresponding frequency of pixel is carried out into gaussian filtering process, can in removable original image
The halation phenomenon that can occur, additionally it is possible to lift the details of image so that image detail is more prominent.
The embodiment of the present invention, due to reference to three in same position in the 3 width high frequency imagings obtained with single order decomposition
The corresponding high-frequency information of pixel calculates the corresponding gain of pixel, therefore, the result for obtaining is that same single order decomposes 3 width for obtaining
The corresponding gain of three pixels in same position is identical in high frequency imaging, i.e., the 3 panel heights frequency for obtaining is decomposed with single order and is schemed
As the same gain diagram of correspondence, so so as to without individually calculating the corresponding gain diagram of every width high frequency imaging, amount of calculation is relatively fewer.
Using the embodiment of the present invention, with the pixel in same position in the 3 width high-frequency decomposition images that single order decomposition is obtained
The corresponding gain of point is identical.
It is further preferred that for each pixel in every width exploded view picture, the pixel is calculated using equation below
Corresponding gain, specifically includes：
Wherein, i represents decomposition exponent number, and j represents that i ranks decompose any width exploded view picture for obtaining, Gij(x, y) represents i ranks point
The exploded view that solution is obtained is as the corresponding gain of pixel in (x, y) position, A in jag(x, y) represents exploded view as being in j
Average frequency value, m after the corresponding filtering of pixel of (x, y) positioniRepresent that i ranks decompose corresponding gain control parameter and 0≤mi
≤ 1, γ are preset constant and 0≤γ≤1, and δ is default constant and 0≤δ≤1, and A' is exploded view as each pixel pair in j
The average value of the frequency answered, ε is for suppressing the exaggerated noise level relevant parameter of noise and 0≤ε≤1, Ak(x, y) is represented
Carry out in the kth width image in the 3N+1 width images obtained after N rank decomposition after the corresponding filtering of pixel in (x, y) position
Frequency.Wherein, the different pixel in (x, y) position, i.e., the pixel with same position in different exploded view pictures.
Specifically, corresponding gain control parameter is decomposed per rank can be the same or different, specific gain control ginseng
Several values can set according to practical application scene, not limit here, it is preferable that same order does not decompose corresponding gain control parameter
It is different.
Wherein, for the different pixel in same exploded view picture, A is worked asagWhen (x, y) is with ε's and more than δ A', i.e. when
AagWhen (x, y) subtracts ε more than δ A', fractionLess than 1, now, Gij(x, y) reduces；Work as Aag(x, y) and ε
And during less than δ A', i.e. work as AijWhen (x, y) subtracts ε less than δ A', fractionMore than 1, now, Gij(x,
Y) increase.
During specific implementation, for each pixel in every width exploded view picture, also the pixel can be calculated using equation below
The corresponding gain of point, specifically includes：
Wherein, i represents decomposition exponent number, and j represents that i ranks decompose any width exploded view picture for obtaining, Gij(x, y) represents i ranks point
The exploded view that solution is obtained is as the corresponding gain of pixel in (x, y) position, A in jij(x, y) is represented in exploded view as in j
Frequency, m after the corresponding filtering of pixel of (x, y) positioniRepresent that i ranks decompose corresponding gain control parameter and 0≤mi≤ 1, γ
It is preset constant and 0≤γ≤1, δ is default constant and 0≤δ≤1, and A' is exploded view as the corresponding frequency of each pixel in j
The average value of rate, ε is for suppressing the exaggerated noise level relevant parameter of noise and 0≤ε≤1.
Specifically, corresponding gain control parameter is decomposed per rank can be the same or different, specific gain control ginseng
Several values can set according to practical application scene, not limit here, it is preferable that same order does not decompose corresponding gain control parameter
It is different.
Wherein, for the different pixel in same exploded view picture, A is worked asijWhen (x, y) is with ε's and more than δ A', i.e. when
AijWhen (x, y) subtracts ε more than δ A', fractionLess than 1, now, Gij(x, y) reduces；Work as Aij(x, y) and ε
And during less than δ A', i.e. work as AijWhen (x, y) subtracts ε less than δ A', fractionMore than 1, now, Gij(x,y)
Increase.
Preferably, the content for being provided according to Fig. 6, determines image after the corresponding treatment of details of the original image：
Step 601, for each the original image vegetarian refreshments in the original image, determines the original image vegetarian refreshments in original image
In position.
Step 602, according to picture corresponding with the position of the original image vegetarian refreshments in original image in original brightness image
Picture corresponding with the position of the original image vegetarian refreshments in original image in luminance picture after the original brightness of vegetarian refreshments, and adjustment
Brightness after the adjustment of vegetarian refreshments, is adjusted to the original pixels point.
Wherein, elaborating on the corresponding pixel in the position position with the original image vegetarian refreshments in original image
The elaboration of corresponding for the position with the pixel in original image pixel in step 402 is referred to, is not done here
Repeat.
Step 603, builds after obtaining the corresponding treatment of original image according to the pixel after each adjustment in original image
Image.
The embodiment of the present invention, each original pixels point for original image carries out treatment of details, so as to reach to original
Image carries out the purpose for the treatment of of details.
Preferably, for each the original image vegetarian refreshments in original image, the original pixels point is carried out using equation below
Treatment of details：
Wherein, Cin(x, y) represents the raw image data of the original image vegetarian refreshments in (x, y) position in original image, Cout
(x, y) represents the view data after being adjusted to the original pixels point in (x, y) position in original image, Iout(x, y) table
Show brightness, I after the adjustment of the pixel in (x, y) position in luminance picture after adjustingin(x, y) is represented in original brightness image
The original brightness of the pixel in (x, y) position.
Based on the inventive concept same with the image processing method that above-described embodiment is provided, the embodiment of the present invention also provides one
Image processing apparatus are planted, as shown in fig. 7, comprises：
Decomposing module 701, for utilizing wavelet transformation, the original brightness image to original image carries out N rank decomposition, and N is
Integer not less than 1；
Frequency regulation block 702, for for every width exploded view picture, according to the width exploded view as corresponding frequency adjusts threshold
Value is adjusted to the corresponding frequency of each pixel in the width exploded view picture, wherein, the frequency more than frequency adjustment threshold value is adjusted
Low, the frequency less than frequency adjustment threshold value is heightened；
Reconstructed module 703, for being adjusted according to frequency after every width exploded view picture, reconstruct the original image corresponding
Luminance picture after adjustment；
Determining module 704, for according to the original image, original brightness image and adjustment after luminance picture, it is determined that
Image after the corresponding treatment of the original image.
Alternatively, in described device, the decomposing module 701, specifically for：
Using Stationary Wavelet Transform, the original brightness image to original image carries out N rank decomposition, obtains 3N width high-frequency decompositions
Image and 1 width low frequency exploded view picture, wherein, being decomposed per rank and obtain 3 width high-frequency decomposition images, last single order is decomposed and also obtains 1
Width low frequency exploded view picture.
Alternatively, in described device, the frequency regulation block 702 is specifically included：
Gain determination unit 7021, for according to the corresponding frequency of each pixel in the width exploded view picture, and the width
Exploded view adjusts threshold value as corresponding frequency, determines the corresponding gain of each pixel in the width exploded view picture；
Frequency adjustment unit 7022, for according to the corresponding gain of each pixel in the width exploded view picture, to the width point
The corresponding frequency of corresponding pixel points is adjusted in solution image.
Alternatively, in described device, the gain determination unit 7021, specifically for：
For each pixel in the width exploded view picture, gaussian filtering process is carried out to the corresponding frequency of the pixel, obtained
Frequency after to the corresponding filtering of the pixel；
Determine each corresponding picture of position in other each width exploded view pictures with the pixel in the width exploded view picture
Frequency after the corresponding filtering of vegetarian refreshments；
Calculate after the corresponding filtering of the pixel in frequency and other each width exploded view pictures with the pixel in the width point
The average value of frequency after the corresponding corresponding filtering of each pixel in position in solution image, as the corresponding filter of the pixel
Average frequency value after ripple；
According to the width exploded view as average frequency value after corresponding frequency adjustment threshold value and the corresponding filtering of the pixel,
Calculate the corresponding gain of the pixel.
Alternatively, in described device, the gain determination unit 7021, specifically for：
If the width exploded view picture is any rank decomposes any high-frequency decomposition image for obtaining, determine that any rank is decomposed
The other two width high-frequency decomposition image for arriving；
Determine the position phase with the pixel in any high-frequency decomposition image in the other two width high frequency imaging
The corresponding frequency of corresponding two pixels；
The absolute value of absolute value and the corresponding frequency of described two pixels according to the corresponding frequency of the pixel, meter
Calculate the corresponding average frequency value of the pixel；
Gaussian filtering process is carried out to the corresponding average frequency value of the pixel, frequency after the corresponding filtering of the pixel is obtained
Rate.
Alternatively, in described device, the gain determination unit 7021 using equation below specifically for calculating the pixel
The corresponding gain of point：
Wherein, i represents decomposition exponent number, and j represents that i ranks decompose any width exploded view picture for obtaining, Gij(x, y) represents i ranks point
The exploded view that solution is obtained is as the corresponding gain of pixel in (x, y) position, A in jag(x, y) represents exploded view as being in j
Average frequency value, m after the corresponding filtering of pixel of (x, y) positioniRepresent that i ranks decompose corresponding gain control parameter and 0≤mi
≤ 1, γ are preset constant and 0≤γ≤1, and δ is default constant and 0≤δ≤1, and A' is exploded view as each pixel pair in j
The average value of the frequency answered, ε is for suppressing the exaggerated noise level relevant parameter of noise and 0≤ε≤1, Ak(x, y) is represented
Carry out in the kth width image in the 3N+1 width images obtained after N rank decomposition after the corresponding filtering of pixel in (x, y) position
Frequency.
Alternatively, in described device, the determining module 704 is specifically included：
Position determination unit 7041, for for each the original image vegetarian refreshments in the original image, determining the original image
Position of the vegetarian refreshments in original image；
Brightness adjustment unit 7042, for according in original brightness image with the position of the original image vegetarian refreshments in original image
Put the original brightness of corresponding pixel, and after adjustment in luminance picture with the position of the original image vegetarian refreshments in original image
Brightness after the adjustment of corresponding pixel is put, the original pixels point is adjusted；
Construction unit 7043, original image correspondence is obtained for being built according to the pixel after each adjustment in original image
Treatment after image.
Alternatively, in described device, the brightness adjustment unit 7042 is specifically for using equation below to the original image
Vegetarian refreshments is adjusted：
Wherein, Cin(x, y) represents the raw image data of the original image vegetarian refreshments in (x, y) position in original image, Cout
(x, y) represents the view data after being adjusted to the original pixels point in (x, y) position in original image, Iout(x, y) table
Show brightness, I after the adjustment of the pixel in (x, y) position in luminance picture after adjustingin(x, y) is represented in original brightness image
The original brightness of the pixel in (x, y) position.
Using image processing method provided in an embodiment of the present invention and device, have the advantages that：By will be original
The luminance picture of image is decomposed, and obtains the exploded view picture comprising frequency, and the frequency to each exploded view picture is adjusted, and
Image Reconstruction is decomposed after being adjusted according to frequency and obtains image after the corresponding adjustment of original image, so as to reach reproduced image details,
And strengthen the purpose of details in the region that details in image is not protruded；According to original brightness image, whole rear luminance picture to original
Beginning image is processed, and so as to obtain image after the corresponding treatment of original image, to improve original image dark for image after the treatment
Place's image detail, and reproduced the details of highlight regions image.
The present invention is the flow with reference to method according to embodiments of the present invention, equipment (system) and computer program product
Figure and/or block diagram are described.It should be understood that every first-class during flow chart and/or block diagram can be realized by computer program instructions
The combination of flow and/or square frame in journey and/or square frame and flow chart and/or block diagram.These computer programs can be provided
The processor of all-purpose computer, special-purpose computer, Embedded Processor or other programmable data processing devices is instructed to produce
A raw machine so that produced for reality by the instruction of computer or the computing device of other programmable data processing devices
The device of the function of being specified in present one flow of flow chart or multiple one square frame of flow and/or block diagram or multiple square frames.
These computer program instructions may be alternatively stored in can guide computer or other programmable data processing devices with spy
In determining the computer-readable memory that mode works so that instruction of the storage in the computer-readable memory is produced and include finger
Make the manufacture of device, the command device realize in one flow of flow chart or multiple one square frame of flow and/or block diagram or
The function of being specified in multiple square frames.
These computer program instructions can be also loaded into computer or other programmable data processing devices so that in meter
Series of operation steps is performed on calculation machine or other programmable devices to produce computer implemented treatment, so as in computer or
The instruction performed on other programmable devices is provided for realizing in one flow of flow chart or multiple flows and/or block diagram one
The step of function of being specified in individual square frame or multiple square frames.
, but those skilled in the art once know basic creation although preferred embodiments of the present invention have been described
Property concept, then can make other change and modification to these embodiments.So, appended claims are intended to be construed to include excellent
Select embodiment and fall into having altered and changing for the scope of the invention.
Obviously, those skilled in the art can carry out various changes and modification without deviating from essence of the invention to the present invention
God and scope.So, if these modifications of the invention and modification belong to the scope of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to comprising these changes and modification.
Claims (16)
1. a kind of image processing method, it is characterised in that including：
Using wavelet transformation, the original brightness image to original image carries out N rank decomposition, and N is the integer not less than 1；
For every width exploded view picture, according to the width exploded view as corresponding frequency adjusts threshold value to each picture in the width exploded view picture
The corresponding frequency of vegetarian refreshments is adjusted, wherein, the frequency more than frequency adjustment threshold value is turned down, less than the frequency that frequency adjusts threshold value
Heighten；
According to the every width exploded view picture after frequency adjustment, luminance picture after the corresponding adjustment of the original image is reconstructed；
According to the original image, original brightness image and luminance picture after adjustment, the corresponding place of the original image is determined
Image after reason.
2. method according to claim 1, it is characterised in that utilize wavelet transformation, the luminance picture to original image enters
Row N ranks are decomposed, and are specifically included：
Using Stationary Wavelet Transform, the original brightness image to original image carries out N rank decomposition, obtains 3N width high-frequency decomposition images
And 1 width low frequency exploded view picture, wherein, being decomposed per rank and obtain 3 width high-frequency decomposition images, last single order decomposes that also to obtain 1 width low
Frequency division solution image.
3. method according to claim 2, it is characterised in that according to the width exploded view as corresponding frequency adjusts threshold value pair
The corresponding frequency of each pixel is adjusted in the width exploded view picture, specifically includes：
According to the corresponding frequency of each pixel in the width exploded view picture, and the width exploded view is as corresponding frequency adjustment threshold
Value, determines the corresponding gain of each pixel in the width exploded view picture；
According to the corresponding gain of each pixel in the width exploded view picture, to the corresponding frequency of corresponding pixel points in the width exploded view picture
Rate is adjusted.
4. method according to claim 3, it is characterised in that determine the corresponding increasing of each pixel in the width exploded view picture
Benefit, specifically includes：
For each pixel in the width exploded view picture, gaussian filtering process is carried out to the corresponding frequency of the pixel, be somebody's turn to do
Frequency after the corresponding filtering of pixel；
Determine each corresponding pixel of position in other each width exploded view pictures with the pixel in the width exploded view picture
Frequency after corresponding filtering；
Calculate after the corresponding filtering of the pixel in frequency and other each width exploded view pictures with the pixel in the width exploded view
The average value of frequency after the corresponding corresponding filtering of each pixel in position as in, after the corresponding filtering of the pixel
Average frequency value；
According to the width exploded view as average frequency value after corresponding frequency adjustment threshold value and the corresponding filtering of the pixel, calculate
The corresponding gain of the pixel.
5. method according to claim 4, it is characterised in that carried out at gaussian filtering to the corresponding frequency of the pixel
Reason, obtains frequency after the corresponding filtering of the pixel, specifically includes：
If the width exploded view picture is any rank decomposes any high-frequency decomposition image for obtaining, determine that any rank decomposes what is obtained
Other two width high-frequency decomposition image；
Determine that the position in the other two width high frequency imaging with the pixel in any high-frequency decomposition image is corresponding
The corresponding frequency of two pixels；
The absolute value of absolute value and the corresponding frequency of described two pixels according to the corresponding frequency of the pixel, calculating should
The corresponding average frequency value of pixel；
Gaussian filtering process is carried out to the corresponding average frequency value of the pixel, frequency after the corresponding filtering of the pixel is obtained.
6. the method according to claim 4 or 5, it is characterised in that the corresponding increasing of the pixel is calculated using equation below
Benefit, specifically includes：
Wherein, i represents decomposition exponent number, and j represents that i ranks decompose any width exploded view picture for obtaining, Gij(x, y) represents that i ranks are decomposed
To exploded view as in j in (x, y) position the corresponding gain of pixel, Aag(x, y) represent exploded view as in j in (x,
Y) average frequency value, m after the corresponding filtering of the pixel of positioniRepresent that i ranks decompose corresponding gain control parameter and 0≤mi≤
1, γ is preset constant and 0≤γ≤1, and δ is default constant and 0≤δ≤1, and A' is exploded view as each pixel correspondence in j
Frequency average value, ε is for suppressing the exaggerated noise level relevant parameter of noise and 0≤ε≤1, Ak(x, y) represent into
Frequency after the corresponding filtering of pixel in (x, y) position in kth width image in the 3N+1 width images that row N ranks obtain after decomposing
Rate.
7. method according to claim 3, it is characterised in that scheme after determining the corresponding treatment of details of the original image
Picture, specifically includes：
For each the original image vegetarian refreshments in the original image, the position of the original image vegetarian refreshments in original image is determined；
According to the original bright of pixel corresponding with the position of the original image vegetarian refreshments in original image in original brightness image
After degree, and adjustment after the adjustment of pixel corresponding with the position of the original image vegetarian refreshments in original image in luminance picture
Brightness, is adjusted to the original pixels point；
Built according to the pixel after each adjustment in original image and obtain image after the corresponding treatment of original image.
8. method according to claim 7, it is characterised in that be adjusted to the original pixels point using equation below：
Wherein, Cin(x, y) represents the raw image data of the original image vegetarian refreshments in (x, y) position in original image, Cout(x,
Y) view data after being adjusted to the original pixels point in (x, y) position in original image, I are representedout(x, y) is represented
Brightness, I after the adjustment of the pixel in (x, y) position in luminance picture after adjustmentin(x, y) locates in representing original brightness image
In the original brightness of the pixel of (x, y) position.
9. a kind of image processing apparatus, it is characterised in that including：
Decomposing module, for utilizing wavelet transformation, the original brightness image to original image carries out N rank decomposition, and N is not less than 1
Integer；
Frequency regulation block, for for every width exploded view picture, according to the width exploded view as corresponding frequency adjusts threshold value to this
The corresponding frequency of each pixel is adjusted in width exploded view picture, wherein, the frequency more than frequency adjustment threshold value is turned down, and is less than
The frequency of frequency adjustment threshold value is heightened；
Reconstructed module, for being adjusted according to frequency after every width exploded view picture, after reconstructing the corresponding adjustment of the original image
Luminance picture；
Determining module, for according to the original image, original brightness image and luminance picture after adjustment, determining described original
Image after the corresponding treatment of image.
10. device according to claim 9, it is characterised in that the decomposing module, specifically for：
Using Stationary Wavelet Transform, the original brightness image to original image carries out N rank decomposition, obtains 3N width high-frequency decomposition images
And 1 width low frequency exploded view picture, wherein, being decomposed per rank and obtain 3 width high-frequency decomposition images, last single order decomposes that also to obtain 1 width low
Frequency division solution image.
11. devices according to claim 10, it is characterised in that the frequency regulation block, specifically include：
Gain determination unit, for according to the corresponding frequency of each pixel in the width exploded view picture, and the width exploded view picture
Corresponding frequency adjusts threshold value, determines the corresponding gain of each pixel in the width exploded view picture；
Frequency adjustment unit, for according to the corresponding gain of each pixel in the width exploded view picture, in the width exploded view picture
The corresponding frequency of corresponding pixel points is adjusted.
12. devices according to claim 11, it is characterised in that the gain determination unit, specifically for：
For each pixel in the width exploded view picture, gaussian filtering process is carried out to the corresponding frequency of the pixel, be somebody's turn to do
Frequency after the corresponding filtering of pixel；
Determine each corresponding pixel of position in other each width exploded view pictures with the pixel in the width exploded view picture
Frequency after corresponding filtering；
Calculate after the corresponding filtering of the pixel in frequency and other each width exploded view pictures with the pixel in the width exploded view
The average value of frequency after the corresponding corresponding filtering of each pixel in position as in, after the corresponding filtering of the pixel
Average frequency value；
According to the width exploded view as average frequency value after corresponding frequency adjustment threshold value and the corresponding filtering of the pixel, calculate
The corresponding gain of the pixel.
13. devices according to claim 12, it is characterised in that the gain determination unit, specifically for：
If the width exploded view picture is any rank decomposes any high-frequency decomposition image for obtaining, determine that any rank decomposes what is obtained
Other two width high-frequency decomposition image；
Determine that the position in the other two width high frequency imaging with the pixel in any high-frequency decomposition image is corresponding
The corresponding frequency of two pixels；
The absolute value of absolute value and the corresponding frequency of described two pixels according to the corresponding frequency of the pixel, calculating should
The corresponding average frequency value of pixel；
Gaussian filtering process is carried out to the corresponding average frequency value of the pixel, frequency after the corresponding filtering of the pixel is obtained.
14. device according to claim 12 or 13, it is characterised in that the gain determination unit is specifically for using such as
Lower formula calculates the corresponding gain of the pixel：
Wherein, i represents decomposition exponent number, and j represents that i ranks decompose any width exploded view picture for obtaining, Gij(x, y) represents that i ranks are decomposed
To exploded view as in j in (x, y) position the corresponding gain of pixel, Aag(x, y) represent exploded view as in j in (x,
Y) average frequency value, m after the corresponding filtering of the pixel of positioniRepresent that i ranks decompose corresponding gain control parameter and 0≤mi≤
1, γ is preset constant and 0≤γ≤1, and δ is default constant and 0≤δ≤1, and A' is exploded view as each pixel correspondence in j
Frequency average value, ε is for suppressing the exaggerated noise level relevant parameter of noise and 0≤ε≤1, Ak(x, y) represent into
Frequency after the corresponding filtering of pixel in (x, y) position in kth width image in the 3N+1 width images that row N ranks obtain after decomposing
Rate.
15. devices according to claim 11, it is characterised in that the determining module, specifically include：
Position determination unit, for for each the original image vegetarian refreshments in the original image, determining the original image vegetarian refreshments in original
Position in beginning image；
Brightness adjustment unit, for according to corresponding with the position of the original image vegetarian refreshments in original image in original brightness image
Pixel original brightness, and after adjustment it is corresponding with the position of the original image vegetarian refreshments in original image in luminance picture
Pixel adjustment after brightness, the original pixels point is adjusted；
Construction unit, after obtaining the corresponding treatment of original image according to the pixel structure after each adjustment in original image
Image.
16. devices according to claim 15, it is characterised in that the brightness adjustment unit is specifically for using following public
Formula is adjusted to the original pixels point：
Wherein, Cin(x, y) represents the raw image data of the original image vegetarian refreshments in (x, y) position in original image, Cout(x,
Y) view data after being adjusted to the original pixels point in (x, y) position in original image, I are representedout(x, y) is represented
Brightness, I after the adjustment of the pixel in (x, y) position in luminance picture after adjustmentin(x, y) locates in representing original brightness image
In the original brightness of the pixel of (x, y) position.
Priority Applications (5)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710021180.5A CN106780400B (en) | 2017-01-11 | 2017-01-11 | Image processing method and device |
| EP17814700.5A EP3459043A4 (en) | 2016-06-21 | 2017-06-20 | Systems and methods for image processing |
| PCT/CN2017/089192 WO2017219962A1 (en) | 2016-06-21 | 2017-06-20 | Systems and methods for image processing |
| US16/219,907 US11094045B2 (en) | 2016-06-21 | 2018-12-13 | Systems and methods for image processing |
| US17/342,695 US20210295480A1 (en) | 2016-06-21 | 2021-06-09 | Systems and methods for image processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710021180.5A CN106780400B (en) | 2017-01-11 | 2017-01-11 | Image processing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106780400A true CN106780400A (en) | 2017-05-31 |
| CN106780400B CN106780400B (en) | 2020-08-04 |
Family
ID=58947807
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710021180.5A Active CN106780400B (en) | 2016-06-21 | 2017-01-11 | Image processing method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106780400B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017219962A1 (en) * | 2016-06-21 | 2017-12-28 | Zhejiang Dahua Technology Co., Ltd. | Systems and methods for image processing |
| CN109034118A (en) * | 2018-08-23 | 2018-12-18 | 吴伟锋 | Real-time wheel body boosting method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1416644A (en) * | 2000-11-09 | 2003-05-07 | 皇家菲利浦电子有限公司 | Method and system for limiting repeatitive presentations based on content |
| US20140112595A1 (en) * | 2012-10-24 | 2014-04-24 | Marvell World Trade Ltd. | Low-frequency compression of high dynamic range images |
| CN104094313A (en) * | 2012-02-01 | 2014-10-08 | 佳能株式会社 | Edge adaptive gain adjustment in image enhancement by frequency decomposition |
| CN104504721A (en) * | 2015-01-08 | 2015-04-08 | 中国科学院合肥物质科学研究院 | Unstructured road detecting method based on Gabor wavelet transformation texture description |
-
2017
- 2017-01-11 CN CN201710021180.5A patent/CN106780400B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1416644A (en) * | 2000-11-09 | 2003-05-07 | 皇家菲利浦电子有限公司 | Method and system for limiting repeatitive presentations based on content |
| CN104094313A (en) * | 2012-02-01 | 2014-10-08 | 佳能株式会社 | Edge adaptive gain adjustment in image enhancement by frequency decomposition |
| US20140112595A1 (en) * | 2012-10-24 | 2014-04-24 | Marvell World Trade Ltd. | Low-frequency compression of high dynamic range images |
| US9324137B2 (en) * | 2012-10-24 | 2016-04-26 | Marvell World Trade Ltd. | Low-frequency compression of high dynamic range images |
| CN104504721A (en) * | 2015-01-08 | 2015-04-08 | 中国科学院合肥物质科学研究院 | Unstructured road detecting method based on Gabor wavelet transformation texture description |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017219962A1 (en) * | 2016-06-21 | 2017-12-28 | Zhejiang Dahua Technology Co., Ltd. | Systems and methods for image processing |
| US11094045B2 (en) | 2016-06-21 | 2021-08-17 | Zhejiang Dahua Technology Co., Ltd. | Systems and methods for image processing |
| CN109034118A (en) * | 2018-08-23 | 2018-12-18 | 吴伟锋 | Real-time wheel body boosting method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106780400B (en) | 2020-08-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3087730B1 (en) | Method for inverse tone mapping of an image | |
| US9135681B2 (en) | Image chroma noise reduction | |
| Lee et al. | A space-variant luminance map based color image enhancement | |
| EP1840830B1 (en) | Image processing method | |
| US8391598B2 (en) | Methods for performing local tone mapping | |
| CN106846276B (en) | Image enhancement method and device | |
| EP1652372B1 (en) | Method of and device for modulating a dynamic range of still and video images | |
| US8160387B2 (en) | Image processing method and computer software for image processing | |
| US20040096103A1 (en) | Method of spatially filtering a digital image using chrominance information | |
| CN106780417A (en) | A kind of Enhancement Method and system of uneven illumination image | |
| TWI387312B (en) | Image noise reduction method and processor | |
| WO2010100609A1 (en) | Method for converting input image data into output image data, image conversion unit for converting input image data into output image data, image processing apparatus, display device | |
| US20040264799A1 (en) | Method of processing an image to form an image pyramid | |
| RU2298223C2 (en) | System and method for correcting dark tones on digital photographs | |
| CN106570831A (en) | Gray image contrast equalization enhancement method | |
| KR20170033395A (en) | Image processing apparatus, image processing method, recording medium, and program | |
| CN110796626A (en) | Image sharpening method and device | |
| JP2007507809A (en) | How to speed up Retinex-type algorithms | |
| KR101134463B1 (en) | Color restoration Retinex method | |
| CN105225205B (en) | Image enchancing method, apparatus and system | |
| CN112019762A (en) | Video processing method and device, storage medium and electronic equipment | |
| CN112288652A (en) | PSO optimization-based guide filtering-Retinex low-illumination image enhancement method | |
| CN106780400A (en) | A kind of image processing method and device | |
| US8565549B2 (en) | Image contrast enhancement | |
| CN108961179B (en) | Medical image post-processing system and using method thereof |
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 |