CN103049886A - Image texture repair method and system - Google Patents
Image texture repair method and system Download PDFInfo
- Publication number
- CN103049886A CN103049886A CN2011103081505A CN201110308150A CN103049886A CN 103049886 A CN103049886 A CN 103049886A CN 2011103081505 A CN2011103081505 A CN 2011103081505A CN 201110308150 A CN201110308150 A CN 201110308150A CN 103049886 A CN103049886 A CN 103049886A
- Authority
- CN
- China
- Prior art keywords
- pixel
- value
- brightness
- image
- adjusted
- 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
Images
Landscapes
- Image Processing (AREA)
- Facsimile Image Signal Circuits (AREA)
Abstract
The invention discloses an image texture repair method, which comprises the following steps of: firstly determining a to-be-adjusted area which requires for texture repair in an image; then filtering the to-be-adjusted area by adopting a filtering method capable of enabling image brightness distribution and texture distribution to be consistent, so as to obtain the brightness value of each pixel in the to-be-adjusted area after filtration; and finally according to the gradient mode of each pixel in the to-be-adjusted area, conducting weighted fusion to the brightness value of each pixel in the to-be-adjusted area after filtration and the original brightness value to obtain the repaired image. The invention additionally discloses an image texture repair system corresponding to the method, another image texture repair method and another image texture repair system. The methods and the systems provided by the invention have the advantages that the repair of image textures can be well realized, and compared with a manual repair method, the repair efficiency of the image textures is greatly improved.
Description
Technical field
The invention belongs to technical field of image processing, be specifically related to a kind of image texture restorative procedure and system.
Background technology
" class " is the problem that often occurs in the printed matter image." class " zone is the single outstanding element within the unit pattern, and is not obvious less than the large figure of amalgamation the time, puts into line but assemble behind the large figure of amalgamation or continuous printing, causes the whole inaesthetic zone of picture." class " shows as the inconsistency of image brightness distribution and grain distribution, but this inconsistency need to be under the guidance of globality, by the careful adjustment of part at image local and not obvious, when finally reaching local natural transition, the effect that overall distribution is consistent.
In the prior art, usually adopt constituency in the Photoshop software, curve adjustment, the function such as fuzzy, artificial continuous switching by entire and part, under the guidance of integral body, once and again trickle adjustment is carried out in the part, finally obtain all very consistent images that does not have " class " of Luminance Distribution and grain distribution.Because above-mentioned manually-operated requires operating personnel to content and the well understanding and grasping of texture of image, therefore need veteran operating personnel to finish.Artificial " class " image of repairing, workload is huge, easily in " my god ".
At present, the method record that does not still have document to repair for image " class " problem does not specially have the application software that designs for image " class " problem specially yet in the practice.
Summary of the invention
For the defective that exists in the prior art, technical matters to be solved by this invention provides a kind of efficient high, effective image texture restorative procedure and system.
For solving the problems of the technologies described above, the technical solution used in the present invention is as follows:
A kind of image texture restorative procedure may further comprise the steps:
Determine to need in the image to carry out the zone to be adjusted that texture is repaired;
Employing can make the image brightness distribution filtering method consistent with grain distribution that filtering is carried out in described zone to be adjusted, obtains the filtered brightness value of each pixel in the described zone to be adjusted;
The filtered brightness value of each pixel and original brightness value in the described zone to be adjusted are weighted fusion, the image after obtaining repairing.
Aforesaid image texture restorative procedure, wherein, the process of filtering may further comprise the steps:
Determine sample areas, and calculate target mean and the target variance of described sample areas;
Travel through all pixels in the described zone to be adjusted, calculate neighboring mean value and the neighborhood variance of each pixel; The span of the described radius of neighbourhood is between 30 pixels～500 pixels;
According to target mean and the target variance of described sample areas, and original brightness value, neighboring mean value and the neighborhood variance of each pixel in the described zone to be adjusted, the filtered brightness value of each pixel in the described zone to be adjusted calculated.
Aforesaid image texture restorative procedure, wherein, adopt following formula to calculate the filtered brightness value of each pixel in the described zone to be adjusted:
Wherein, I
w(i, j) represents filtered brightness value, M
dAnd S
dRepresent respectively target mean and target variance, I (i, j) is illustrated in the original brightness value that (i, j) locates pixel, and M (i, j) and S (i, j) represent that respectively (i, j) locates neighboring mean value and the neighborhood variance of pixel.
Aforesaid image texture restorative procedure, if there is dark side in described image, then described method is adjusted to the overall brightness distribution of appointed area in the image consistent with the overall brightness distribution of sample areas first.The process of described adjustment may further comprise the steps:
Need to carry out zone and the sample areas of brightness adjustment in the specify image, the zone that needs in the image to carry out brightness adjustment is designated as the appointed area;
Calculate respectively the histogram of described sample areas and appointed area, the column hisgram of going forward side by side coupling obtains the brightness mapping relations from described appointed area to described sample areas;
Described brightness mapping relations are applied to each pixel in the described appointed area.
Aforesaid image texture restorative procedure, wherein, the process of Weighted Fusion may further comprise the steps:
Calculate the gradient-norm value of each pixel in the described zone to be adjusted, and calculate the Sigmoid functional value about described gradient-norm value, described Sigmoid function is as follows:
Wherein, || G (i, j) || be independent variable, (i, j) locates the gradient-norm of pixel on the presentation video; α and β are constant, and be preferred, α=-0.15, and pace of change and the degrees of offset of Sigmoid curve are controlled respectively in β=-10;
To each pixel in the described zone to be adjusted, carry out the Weighted Fusion about original brightness value and filtered brightness value take described Sigmoid functional value as weight.Fusion formula is as follows:
I′(i，j)＝ζ(||G(i，j)||)×I
w(i，j)+(1-ζ(||G(i，j)||))×I(i，j)
Wherein, (i, j) located the end value of pixel, I after I ' (i, j) expression was merged
wThe filtered brightness value of pixel is located in (i, j) expression (i, j), and the original brightness value of pixel is located in I (i, j) expression (i, j).
A kind of image texture restorative procedure may further comprise the steps:
Determine to need in the image to carry out zone to be adjusted and the sample areas that texture is repaired;
Calculate the histogram of described sample areas;
Travel through all pixels in the described zone to be adjusted, each pixel is handled as follows: the 1. histogram of calculating pixel vertex neighborhood, 2. according to histogram and the histogram calculation neighborhood of pixel points of neighborhood and the brightness mapping relations of described sample areas of described sample areas, 3. calculate the brightness mapping value of all pixels in the neighborhood according to described brightness mapping relations; Preferably, the span of the described radius of neighbourhood is between 20 pixels～100 pixels;
After traversal finished, one group of brightness mapping value that each pixel is tried to achieve was weighted on average, obtains the final brightness mapping value of this pixel;
Final brightness mapping value and original brightness value to each pixel in the described zone to be adjusted are weighted fusion, the image after obtaining repairing.
Aforesaid image texture restorative procedure, wherein, the process of Weighted Fusion may further comprise the steps:
Calculate the gradient-norm value of each pixel in the described zone to be adjusted, and calculate the Sigmoid functional value about described gradient-norm value, described Sigmoid function is as follows:
Wherein, || G (i, j) || be independent variable, (i, j) locates the gradient-norm of pixel on the presentation video; α and β are constant, α=-0.15, and pace of change and the degrees of offset of Sigmoid curve are controlled respectively in β=-30;
To each pixel in the described zone to be adjusted, carry out the Weighted Fusion about original brightness value and described final brightness mapping value take described Sigmoid functional value as weight.Fusion formula is as follows:
I′(i，j)＝ζ(||G(i，j)||)×I
w(i，j)+(1-ζ(||G(i，j)||))×(i，j)
Wherein, (i, j) located the end value of pixel, I after I ' (i, j) expression was merged
w(i, j) expression step 4. in (i, j) locate the final brightness value of pixel, the original brightness value of pixel is located in I (i, j) expression (i, j).
A kind of image texture repair system comprises for determining that image needs to carry out definite device in the zone to be adjusted that texture repairs;
Be used for to adopt to make the image brightness distribution filtering method consistent with grain distribution treat adjustment region to carry out filtering, obtain in the zone to be adjusted the filter of brightness value after each pixel filtering;
Be used for treating the fusing device that the filtered brightness value of each pixel and original brightness value in the adjustment region are weighted fusion.
Aforesaid image texture repair system, wherein, filter comprises for the determining unit of determining sample areas;
Be used for traveling through the traversal unit of all pixels in the zone to be adjusted;
Be used for calculating target mean and the target variance of sample areas, and calculate the neighboring mean value of each pixel and the first computing unit of neighborhood variance;
Be used for target mean and target variance according to sample areas, and original brightness value, neighboring mean value and the neighborhood variance of each pixel in the zone to be adjusted, calculate the second computing unit of the filtered brightness value of each pixel in the zone to be adjusted.
Aforesaid image texture repair system, also comprising distributes the overall brightness in described zone to be adjusted is adjusted to and the overall brightness of the sample areas of the appointment consistent brightness adjusting device that distributes.
Aforesaid image texture repair system, wherein, brightness adjusting device comprises being used to specify needs to carry out the zone of brightness adjustment and the designating unit of sample areas in the image, the zone that needs in the image to carry out brightness adjustment is designated as the appointed area;
Be used for calculating sample areas and histogrammic the 3rd computing unit in appointed area;
Be used for coupling appointed area histogram and sample areas histogram, obtain the matching unit of the brightness mapping relations from the appointed area to the sample areas;
Be used for the brightness mapping relations are applied to the applying unit of each pixel in the appointed area.
Aforesaid image texture repair system, wherein, fusing device comprises be used to the gradient-norm value of calculating each pixel in the described zone to be adjusted, and calculating is about the 4th computing unit of the Sigmoid functional value of described gradient-norm value;
Be used for treating each pixel in the adjustment region, carry out the integrated unit about the Weighted Fusion of brightness value after original brightness value and the filtering take the Sigmoid functional value as weight.
A kind of image texture repair system comprises for determining that image needs to carry out the zone to be adjusted of texture reparation and definite device of sample areas;
Be used for calculating sample areas histogram and histogrammic the first calculation element of neighborhood of pixel points;
Be used for traveling through the traversal device of all pixels in the zone to be adjusted;
Be used for the brightness mapping relations according to described sample areas histogram and neighborhood histogram calculation neighborhood of pixel points and sample areas, calculate all pixel intensity mapping value in the neighborhood according to the brightness mapping relations, and after traversal finishes, one group of brightness mapping value that each pixel is tried to achieve is weighted on average, obtains the second calculation element of the final brightness mapping value of this pixel;
Be used for treating the final brightness mapping value of each pixel in the adjustment region and the fusing device that the original brightness value is weighted fusion.
Aforesaid image texture repair system, wherein, fusing device comprises be used to the gradient-norm value of calculating each pixel in the described zone to be adjusted, and calculating is about the computing unit of the Sigmoid functional value of described gradient-norm value;
Be used for treating each pixel in the adjustment region, carry out the integrated unit about the Weighted Fusion of original brightness value and described final brightness mapping value take the Sigmoid functional value as weight.
The method of the invention and system, the methods such as Wallis filtering, local average Histogram Matching have been incorporated into class image repair field, by image being carried out after the modes such as Wallis filtering, local average Histogram Matching adjust, be weighted fusion take the Sigmoid functional value as weight to adjusting rear image and original image again, realized well the reparation of image texture.And, compare with artificial repair mode, greatly improved the remediation efficiency of image texture.Test the speed through experiment, under Pentium double-core, 3.46G memory environment, to one 10000 * 10000 image, the processing time is substantially in 1 minute.
Description of drawings
Fig. 1 is the structured flowchart of image texture repair system in the embodiment 1;
Fig. 2 is the structured flowchart of brightness adjusting device in the embodiment 1;
Fig. 3 is the structured flowchart of filter in the embodiment 1;
Fig. 4 is the structured flowchart of fusing device in the embodiment 1;
Fig. 5 is the process flow diagram of image texture restorative procedure in the embodiment 1;
Fig. 6 is the process flow diagram of luminance regulating method in the embodiment 1;
Fig. 7 is the process flow diagram of Wallis filtering method in the embodiment 1;
Fig. 8 is Sigmoid curve synoptic diagram in the embodiment 1;
Fig. 9 is the structured flowchart of image texture repair system in the embodiment 2;
Figure 10 is the structured flowchart of fusing device in the embodiment 2;
Figure 11 is the process flow diagram of image texture restorative procedure in the embodiment 2.
Embodiment
The present invention be directed to the part class problem that exists in the image and the solution that proposes, its core concept is: after the modes such as filtering, local average Histogram Matching that image is carried out are adjusted, be weighted fusion take the Sigmoid functional value as weight to adjusting rear image and original image again, thereby realize repairing preferably the purpose of image texture.Below in conjunction with accompanying drawing the specific embodiment of the present invention is elaborated.
Embodiment 1
Class is a class problem that often occurs in the printed matter image, even image brightness distribution is unified, still may exist local unit to assemble line through connecting after solarizations waits operation, and the not problem in zone attractive in appearance is given prominence in formation.For this problem, need to carry out part adjustment to image, consistent with the grain distribution of realizing entire image.But to containing the image of large structural texture, when realizing that grain distribution is consistent, often need to keep well the structure of image.Present embodiment provides a kind of image texture repair system and method for above-mentioned image class problem.
As shown in Figure 1, in the present embodiment image texture repair system comprise brightness adjusting device 15, determine device 11, filter 12 and fusing device 13.As shown in Figure 2, brightness adjusting device 15 comprises designating unit 151, the 3rd computing unit 152, matching unit 153 and applying unit 154.As shown in Figure 3, filter 12 comprises determining unit 121, traversal unit 122, the first computing unit 123 and the second computing unit 124.As shown in Figure 4, fusing device 13 comprises the 4th computing unit 131 and integrated unit 132.
Determine that device 11 is used for determining that image needs to carry out the zone to be adjusted that texture is repaired.
Fusing device 13 is used for treating the interior filtered brightness value of each pixel of adjustment region and the original brightness value is weighted fusion.Wherein: the 4th computing unit 131 is used for calculating the gradient-norm value of each pixel in the described zone to be adjusted, and calculates the Sigmoid functional value about described gradient-norm value; Integrated unit 132 is used for treating each pixel in the adjustment region, carries out the Weighted Fusion about original brightness value and filtered brightness value take the Sigmoid functional value as weight.
As shown in Figure 5, the method that adopts said system to repair image texture may further comprise the steps:
(1) brightness adjusting device 15 is adjusted to the overall brightness distribution of appointed area in the image consistent with the overall brightness distribution of sample areas.
Brightness adjustment is the solution that proposes for the dark side problem in the image class problem, if therefore do not have dark side in the image, then this step can be omitted.Dark side out-of-flatness during owing to the splicing of scintigram in batches, the scanning of original copy stock, scanning thing material itself have the reason such as reflecting effect to cause.The dark side that exists in the image has the advantages that bulk distributes, and therefore is easy to manually choose out the image-region with consistent brightness.Based on These characteristics, the purpose of brightness adjustment operation is in this step: the Luminance Distribution in the zone of user's appointment in the image is adjusted to the consistent effect of sample areas that provides with the user.
As shown in Figure 6, the process of brightness adjustment may further comprise the steps:
1. need to carry out zone and the sample areas of brightness adjustment in designating unit 151 specify images, the zone that needs in the image to carry out brightness adjustment is designated as the appointed area.
Target effect for reference when sample areas herein refers to carry out brightness adjustment is regional, can be the regional area in the zone to be adjusted, also can be extra-regional other image-region to be adjusted in the image, can also be other outer image-regions of this image, need before processing, to pre-determine.
2. the 3rd computing unit 152 calculates respectively sample areas histogram and appointed area histogram, and matching unit 153 carries out Histogram Matching, obtains the brightness mapping relations from the appointed area to the sample areas.
3. applying unit 154 is applied to each pixel in the appointed area with the brightness mapping relations, obtains the Luminance Distribution image consistent with sample areas.
(2) determine to need to carry out the zone to be adjusted that texture is repaired in device 11 definite images.
Because class is a class problem that often occurs in the printed matter image, even if the brightness adjustment mode in the employing step (1) has realized the unification of integral image Luminance Distribution, assemble line after still may existing local unit to pass through operations such as connecting solarization, form the outstanding not problem in zone attractive in appearance.For this problem, need to carry out part adjustment to image, consistent with the grain distribution of realizing entire image.
(3) filter 12 adopts and can make the image brightness distribution filtering method consistent with grain distribution that filtering is carried out in described zone to be adjusted, obtains the filtered brightness value of each pixel in the described zone to be adjusted.
For the image that contains large structural texture, when usually wishing to realize that grain distribution is consistent, keep well the structure of image.For this reason, present embodiment adopts the Wallis filtering method to treat adjustment region to carry out filtering.
Wallis filtering is a kind of highly effective realization image brightness distribution filtering method consistent with grain distribution, and fundamental formular is as follows:
Wherein, I
w(i, j) represents filtered brightness value, M
dAnd S
dRepresent respectively target mean and target variance, I (i, j) is illustrated in the original brightness value that (i, j) locates pixel, and M (i, j) and S (i, j) represent that respectively (i, j) locates neighboring mean value and the neighborhood variance of pixel.
According to above-mentioned formula, each pixel neighborhood of a point has approximate consistent average M on the image after the adjustment
dWith variance S
d, same average has realized the brightness uniformity of entire image, and it is consistent that same variance has realized that image texture distributes, and the selection of neighborhood scope has determined the maintenance degree of picture structure.
As shown in Figure 7, concrete filtering may further comprise the steps:
1. determining unit 121 is determined sample areas, and the first computing unit 123 calculates average and the variance of sample areas, and the average of sample areas is called target mean, and the variance of sample areas is called the target variance.
Target effect for reference when sample areas herein refers to carry out filtering is regional, determined by determining unit 121 by the user, the sample areas of using from above-mentioned brightness adjustment can be different, but can be the regional area in the zone to be adjusted equally, also can be extra-regional other image-region to be adjusted in the image, can also be other outer image-regions of this image, need before processing, to pre-determine.
2. travel through all pixels in the 122 traversals zone to be adjusted, unit, the first computing unit 123 calculates neighboring mean value and the neighborhood variance of each pixel.
The selection of the radius of neighbourhood is great on the impact of filter effect, if wish that picture structure is average as far as possible, and need not keep picture structure, then can adopt less radius to carry out filtering.The effective range of the general radius of neighbourhood is between 30 pixels～500 pixels, and the too little meeting of radius causes the structure loss of image serious, and radius too greatly then makes the adjustment DeGrain.In the present embodiment, the radius of neighbourhood is got 100 pixels, and namely neighborhood is 201 * 201 square, to guarantee the maintenance to picture structure.
3. the second computing unit 124 is according to target mean and the target variance of sample areas, and original brightness value, neighboring mean value and the neighborhood variance of each pixel in the zone to be adjusted, adopt above-mentioned Wallis Filtering Formula to calculate the filtered brightness value of each pixel in the zone to be adjusted.
(4) fusing device 13 treats that the filtered brightness value of each pixel and original brightness value are weighted fusion in the adjustment region.
Adopt original Wallis filtering method can realize that image texture and brightness consistance on the whole distributes, but have following two problems:
1. according to the texture feature of sample areas, picture contrast may be needed to increase, and to the original regional area that just has larger contrast on the image, when further widening contrast, Patch effect may be occurred.
The formation of patch, be since in the image brightness and the mean flow rate in certain piece zone differ greatly, i.e. local very bright or dark zone increases or subtracts according to target average brightness again and can obtain the same brightness value that equates with the maximum value or minimum value of brightness range because blocking above brightness range this value.In addition, this zone is the stabilized zone close with neighborhood, thereby the adjusted value of whole neighborhood all is very bright or very dark value, thereby forms patch.
2. the effect of Wallis filtering is to realize that brightness and the variance of image local are in full accord, and processing procedure is not considered any structural information, thereby may break the architectural feature of changing image.For this reason, be necessary the stabilized zone of token image structure in the image is protected, to realize the effective maintenance to picture structure.
Accordingly, present embodiment adopts image and original image after the Wallis filtering to be weighted the method solution above-mentioned two problems of fusion.
Because patch is a very little part in the image, most end value should be the result of Wallis filtering, therefore needs to adopt the weighting function that can keep as far as possible Wallis filtering result.In the present embodiment, adopt the following Sigmoid weighting function about gradient-norm:
In the following formula, || G (i, j) || be independent variable, (i, j) locates the gradient-norm of pixel on the presentation video; α and β are constant, control respectively pace of change and the degrees of offset of Sigmoid curve.The value of α and β can change according to actual needs, in the present embodiment, (being the flat site in the image) keeps the image original value at the very little place of gradient-norm value in requirement, and the filtering end value is adopted in the larger zone of other gradient-norm values, therefore for when removing patch as far as possible, keeps the result of Wallis filtering, adopt rising very fast, following parameter near the null value weight is arranged near 0: α=-0.15, β=-10, its corresponding Sigmoid curve map is as shown in Figure 8.
In the present embodiment, the process of Weighted Fusion is as follows:
At first the 4th computing unit 131 calculates the gradient-norm value of each pixel in the zone to be adjusted, again according to the Sigmoid functional value of above-mentioned formula calculating about described gradient-norm value.The gradient-norm value of calculating each pixel herein is based on that the pixel value (be original pixel value) of zone to be adjusted before adjustment carry out, because weight is to consider from the stability of image itself.
Then integrated unit 132 is treated each pixel in the adjustment region, carries out the Weighted Fusion about original brightness value and filtered brightness value take the Sigmoid functional value as weight, and its fusion formula is as follows:
I′(i，j)＝ζ(||G(i，j)||)×I
w(i，j)+(1-ζ(||G(i，j)||))×I(i，j)
Wherein, (i, j) located the end value of pixel, I after I ' (i, j) presentation video merged
wThe filtering end value of pixel is located in (i, j) expression (i, j), and the original brightness value of pixel is located in I (i, j) expression (i, j).
Embodiment 2
Some image has the consistance texture, this type of image is easily after the operations such as Image Mosaics, the zone that local grain is sparse or dense is combined into larger zone and forms the Luminance Distribution such as patch and the inconsistent class of grain distribution zone, such as ever-present " hickie " phenomenon that is formed by the sparse texture set in the printed matter image.Present embodiment provides a kind of image texture repair system and method for the class problem for having local luminance or the inconsistent distribution of texture in the consistance texture image.
As shown in Figure 9, the image texture repair system comprises definite device 111, the first calculation element 112, traversal device 113, the second calculation element 114 and fusing device 115 in the present embodiment.As shown in figure 10, fusing device 115 comprises computing unit 1151 and integrated unit 1152.
Determine that device 111 is used for determining that image needs to carry out zone to be adjusted and the sample areas that texture is repaired.The first calculation element 112 is used for calculating sample areas histogram and neighborhood of pixel points histogram.Traversal device 113 is used for traveling through all pixels in the zone to be adjusted.The brightness mapping relations that the second calculation element 114 is used for according to sample areas histogram and neighborhood histogram calculation neighborhood of pixel points and sample areas, calculate all pixel intensity mapping value in the neighborhood according to the brightness mapping relations, and after traversal finishes, one group of brightness mapping value that each pixel is tried to achieve is weighted on average, obtains the final brightness mapping value of this pixel.Fusing device 115 is weighted fusion for final brightness mapping value and the original brightness value for the treatment of each pixel in the adjustment region.Wherein: computing unit 1151 is used for calculating the gradient-norm value of each pixel in the zone to be adjusted, and calculates the Sigmoid functional value about described gradient-norm value; Integrated unit 1152 is used for treating each pixel in the adjustment region, carries out the Weighted Fusion about original brightness value and final brightness mapping value take the Sigmoid functional value as weight.
As shown in figure 11, the method that adopts said system to carry out the image texture reparation may further comprise the steps:
(1) determines to need to carry out zone to be adjusted and the sample areas that texture is repaired in device 111 definite images.Target effect for reference when described sample areas refers to carry out the texture reparation is regional, can be the regional area in the zone to be adjusted, also can be extra-regional other image-region to be adjusted in the image, can also be other outer image-regions of this image, need before processing, to pre-determine.
(2) first calculation elements 112 calculate the histogram of sample areas.
(3) all pixels in the traversal device 113 traversals zone to be adjusted are handled as follows each pixel:
1. the histogram of the first calculation element 112 calculating pixel vertex neighborhoods; The effective range of the general radius of neighbourhood is between 20 pixels～100 pixels, and the too little meeting of radius causes Neighborhood Statistics information inaccurate and to adjust the effect deviation too large, radius is not too greatly then had singularity so that adjust DeGrain because of neighborhood information.In the present embodiment, the radius of neighbourhood is got 50 pixels, and namely neighborhood is 101 * 101 square;
2. the second calculation element 114 is according to histogram and the histogram calculation neighborhood of pixel points of neighborhood and the brightness mapping relations of sample areas of sample areas;
3. calculate the brightness mapping value of all pixels in the neighborhood according to the brightness mapping relations;
(4) after traversal finished, one group of brightness mapping value that the second calculation element 114 is tried to achieve each pixel was weighted on average, obtains the final brightness mapping value of this pixel.
(5) final brightness mapping value and the original brightness value of each pixel is weighted fusion in 115 pairs of described zones to be adjusted of fusing device.
Because the texture of such image is inner and the background area all is the stabilized zone, the stabilized zone need not to adjust again.Therefore, present embodiment adopts the mode based on the Weighted Fusion of image and original image after the adjustment of Sigmoid function to guarantee that the stabilized zone does not adjust.
At first, computing unit 1151 calculates the gradient-norm value of each pixel in the zone to be adjusted, again according to the Sigmoid functional value of above-mentioned formula calculating about described gradient-norm value.Described Sigmoid function is as follows:
In the following formula, || G (i, j) || be independent variable, (i, j) locates the gradient-norm of pixel on the presentation video; α and β are constant, control respectively pace of change and the degrees of offset of Sigmoid curve.Because present embodiment is applied to have in the class image repair of consistance texture usually, it focuses on the adjustment to transitional region in the image, than embodiment 1, when requiring the Sigmoid functional value to change with the gradient-norm value is very fast, variation range is also larger, therefore, α=-0.15, β=-30.
Then, integrated unit 1152 is treated each pixel in the adjustment region, carries out the Weighted Fusion about original brightness value and final brightness mapping value take the Sigmoid functional value as weight, and its fusion formula is as follows:
I′(i，j)＝ζ(||G(i，j)||)×I
w(i，j)+(1-ζ(||G(i，j)||))×I(i，j)
Wherein, (i, j) located the end value of pixel, I after I ' (i, j) presentation video merged
wThe final brightness mapping value of pixel after step (3) is processed located in (i, j) expression (i, j), and (i, j) locates the original brightness value of pixel in I (i, the j) presentation video.
Obviously, those skilled in the art can carry out various changes and modification to the present invention and not break away from the spirit and scope of the present invention.For example, fusion method can adopt any fusion method that meets specific image, and Wallis function and Sigmoid function also can adopt other variants based on basic function.Like this, if of the present invention these are revised and modification belongs within the scope of claim of the present invention and equivalent technology thereof, then the present invention also is intended to comprise these changes and modification interior.
Claims (21)
1. an image texture restorative procedure is characterized in that, may further comprise the steps:
Determine to need in the image to carry out the zone to be adjusted that texture is repaired;
Employing can make the image brightness distribution filtering method consistent with grain distribution that filtering is carried out in described zone to be adjusted, obtains the filtered brightness value of each pixel in the described zone to be adjusted;
The filtered brightness value of each pixel and original brightness value in the described zone to be adjusted are weighted fusion, the image after obtaining repairing.
2. image texture restorative procedure as claimed in claim 1 is characterized in that, the process of described filtering may further comprise the steps:
Determine sample areas, and calculate target mean and the target variance of described sample areas;
Travel through all pixels in the described zone to be adjusted, calculate neighboring mean value and the neighborhood variance of each pixel;
According to target mean and the target variance of described sample areas, and original brightness value, neighboring mean value and the neighborhood variance of each pixel in the described zone to be adjusted, the filtered brightness value of each pixel in the described zone to be adjusted calculated.
3. image texture restorative procedure as claimed in claim 2 is characterized in that, the span of the described radius of neighbourhood is between 30 pixels～500 pixels.
4. image texture restorative procedure as claimed in claim 2 is characterized in that, adopts following formula to calculate the filtered brightness value of each pixel in the described zone to be adjusted:
Wherein, I
w(i, j) represents filtered brightness value, M
dAnd S
dRepresent respectively target mean and target variance, I (i, j) is illustrated in the original brightness value that (i, j) locates pixel, and M (i, j) and S (i, j) represent that respectively (i, j) locates neighboring mean value and the neighborhood variance of pixel.
5. such as each described image texture restorative procedure in the claim 1～4, it is characterized in that: if there is dark side in described image, then described method is adjusted to the overall brightness distribution of appointed area in the image consistent with the overall brightness distribution of sample areas first.
6. image texture restorative procedure as claimed in claim 5 is characterized in that, the process of described adjustment may further comprise the steps:
Need to carry out zone and the sample areas of brightness adjustment in the specify image, the zone that needs in the image to carry out brightness adjustment is designated as the appointed area;
Calculate respectively the histogram of described sample areas and appointed area, the column hisgram of going forward side by side coupling obtains the brightness mapping relations from described appointed area to described sample areas;
Described brightness mapping relations are applied to each pixel in the described appointed area.
7. image texture restorative procedure as claimed in claim 1 is characterized in that, the process of described Weighted Fusion may further comprise the steps:
Calculate the gradient-norm value of each pixel in the described zone to be adjusted, and calculate the Sigmoid functional value about described gradient-norm value, described Sigmoid function is as follows:
Wherein, || G (i, j) || be independent variable, (i, j) locates the gradient-norm of pixel on the presentation video; α and β are constant, control respectively pace of change and the degrees of offset of Sigmoid curve;
To each pixel in the described zone to be adjusted, carry out the Weighted Fusion about original brightness value and filtered brightness value take described Sigmoid functional value as weight.
8. image texture restorative procedure as claimed in claim 7 is characterized in that: described α=-0.15, β=-10.
9. such as claim 7 or 8 described image texture restorative procedures, it is characterized in that the formula of described Weighted Fusion is as follows:
I′(i，j)＝ζ(||G(i，j)||)×I
w(i，j)+(1-ζ(||G(i，j)||))×I(i，j)
Wherein, (i, j) located the end value of pixel, I after I ' (i, j) expression was merged
wThe filtered brightness value of pixel is located in (i, j) expression (i, j), and the original brightness value of pixel is located in I (i, j) expression (i, j).
10. an image texture restorative procedure is characterized in that, may further comprise the steps:
Determine to need in the image to carry out zone to be adjusted and the sample areas that texture is repaired;
Calculate the histogram of described sample areas;
Travel through all pixels in the described zone to be adjusted, each pixel is handled as follows: the 1. histogram of calculating pixel vertex neighborhood, 2. according to histogram and the histogram calculation neighborhood of pixel points of neighborhood and the brightness mapping relations of described sample areas of described sample areas, 3. calculate the brightness mapping value of all pixels in the neighborhood according to described brightness mapping relations;
After traversal finished, one group of brightness mapping value that each pixel is tried to achieve was weighted on average, obtains the final brightness mapping value of this pixel;
Final brightness mapping value and original brightness value to each pixel in the described zone to be adjusted are weighted fusion, the image after obtaining repairing.
11. image texture restorative procedure as claimed in claim 10 is characterized in that, the span of the described radius of neighbourhood is between 20 pixels～100 pixels.
12. such as claim 10 or 11 described image texture restorative procedures, it is characterized in that the process of described Weighted Fusion may further comprise the steps:
Calculate the gradient-norm value of each pixel in the described zone to be adjusted, and calculate the Sigmoid functional value about described gradient-norm value, described Sigmoid function is as follows:
Wherein, || G (i, j) || be independent variable, (i, j) locates the gradient-norm of pixel on the presentation video; α and β are constant, control respectively pace of change and the degrees of offset of Sigmoid curve;
To each pixel in the described zone to be adjusted, carry out the Weighted Fusion about original brightness value and described final brightness mapping value take described Sigmoid functional value as weight.
13. image texture restorative procedure as claimed in claim 12 is characterized in that: described α=-0.15, β=-30.
14. image texture restorative procedure as claimed in claim 12 is characterized in that, described fusion formula is as follows:
I′(i，j)＝ζ(||G(i，j)||)×I
w(i，j)+(1-ζ(||G(i，j)||))×I(i，j)
Wherein, (i, j) located the end value of pixel, I after I ' (i, j) expression was merged
w(i, j) expression step 4. in (i, j) locate the final brightness value of pixel, the original brightness value of pixel is located in I (i, j) expression (i, j).
15. an image texture repair system is characterized in that: comprise for determining that image needs to carry out definite device (11) in the zone to be adjusted that texture repairs;
Be used for to adopt to make the image brightness distribution filtering method consistent with grain distribution treat adjustment region to carry out filtering, obtain in the zone to be adjusted the filter (12) of brightness value after each pixel filtering;
Be used for treating the fusing device (13) that the filtered brightness value of each pixel and original brightness value in the adjustment region are weighted fusion.
16. image texture repair system as claimed in claim 15 is characterized in that: described filter (12) comprises for the determining unit (121) of determining sample areas;
Be used for traveling through the traversal unit (122) of all pixels in the zone to be adjusted;
Be used for calculating target mean and the target variance of sample areas, and calculate the neighboring mean value of each pixel and first computing unit (123) of neighborhood variance;
Be used for target mean and target variance according to sample areas, and original brightness value, neighboring mean value and the neighborhood variance of each pixel in the zone to be adjusted, calculate second computing unit (124) of the filtered brightness value of each pixel in the zone to be adjusted.
17. such as claim 15 or 16 described image texture repair systems, it is characterized in that: described system also comprises the overall brightness in described zone to be adjusted distributed and is adjusted to and the overall brightness of the sample areas of the appointment consistent brightness adjusting device (15) that distributes.
18. image texture repair system as claimed in claim 17, it is characterized in that: described brightness adjusting device (15) comprises being used to specify needs to carry out the zone of brightness adjustment and the designating unit of sample areas (151) in the image, the zone that needs in the image to carry out brightness adjustment is designated as the appointed area;
Be used for calculating sample areas and histogrammic the 3rd computing unit in appointed area (152);
Be used for coupling appointed area histogram and sample areas histogram, obtain the matching unit (153) of the brightness mapping relations from the appointed area to the sample areas;
Be used for the brightness mapping relations are applied to the applying unit (154) of each pixel in the appointed area.
19. such as claim 15 or 16 described image texture repair systems, it is characterized in that: described fusing device (13) comprises be used to the gradient-norm value of calculating each pixel in the described zone to be adjusted, and calculating is about the 4th computing unit (131) of the Sigmoid functional value of described gradient-norm value;
Be used for treating each pixel in the adjustment region, carry out the integrated unit (132) about the Weighted Fusion of brightness value after original brightness value and the filtering take the Sigmoid functional value as weight.
20. an image texture repair system is characterized in that, comprises for determining that image needs to carry out the zone to be adjusted of texture reparation and definite device (111) of sample areas;
Be used for calculating sample areas histogram and histogrammic the first calculation element of neighborhood of pixel points (112);
Be used for traveling through the traversal device (113) of all pixels in the zone to be adjusted;
Be used for the brightness mapping relations according to described sample areas histogram and neighborhood histogram calculation neighborhood of pixel points and sample areas, calculate all pixel intensity mapping value in the neighborhood according to the brightness mapping relations, and after traversal finishes, one group of brightness mapping value that each pixel is tried to achieve is weighted on average, obtains second calculation element (114) of the final brightness mapping value of this pixel;
Be used for treating the final brightness mapping value of each pixel in the adjustment region and the fusing device (115) that the original brightness value is weighted fusion.
21. image texture repair system as claimed in claim 20, it is characterized in that: described fusing device (115) comprises be used to the gradient-norm value of calculating each pixel in the described zone to be adjusted, and calculating is about the computing unit (1151) of the Sigmoid functional value of described gradient-norm value;
Be used for treating each pixel in the adjustment region, carry out the integrated unit (1152) about the Weighted Fusion of original brightness value and described final brightness mapping value take the Sigmoid functional value as weight.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110308150.5A CN103049886B (en) | 2011-10-12 | 2011-10-12 | A kind of image texture repair method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110308150.5A CN103049886B (en) | 2011-10-12 | 2011-10-12 | A kind of image texture repair method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103049886A true CN103049886A (en) | 2013-04-17 |
| CN103049886B CN103049886B (en) | 2015-10-28 |
Family
ID=48062516
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110308150.5A Active CN103049886B (en) | 2011-10-12 | 2011-10-12 | A kind of image texture repair method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103049886B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105260997A (en) * | 2015-09-22 | 2016-01-20 | 北京好运到信息科技有限公司 | Method for automatically obtaining target image |
| CN106650794A (en) * | 2016-11-24 | 2017-05-10 | 北京理工大学 | Method and system for eliminating highlight of image affected by highlight reflection on object surface |
| CN107622481A (en) * | 2017-10-25 | 2018-01-23 | 沈阳东软医疗系统有限公司 | Reduce the method, apparatus and computer equipment of CT picture noises |
| CN108830330A (en) * | 2018-06-22 | 2018-11-16 | 西安电子科技大学 | Classification of Multispectral Images method based on self-adaptive features fusion residual error net |
| CN109816615A (en) * | 2019-03-06 | 2019-05-28 | 腾讯科技（深圳）有限公司 | Image processing method, device, equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1742178A2 (en) * | 2005-07-05 | 2007-01-10 | Xerox Corporation | Contrast enhancement of images |
| CN101569199A (en) * | 2007-10-03 | 2009-10-28 | 松下电器产业株式会社 | Method and device for reducing block distortion |
| CN101770639A (en) * | 2010-01-14 | 2010-07-07 | 北京航空航天大学 | Enhancement method of low-illumination image |
-
2011
- 2011-10-12 CN CN201110308150.5A patent/CN103049886B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1742178A2 (en) * | 2005-07-05 | 2007-01-10 | Xerox Corporation | Contrast enhancement of images |
| CN101569199A (en) * | 2007-10-03 | 2009-10-28 | 松下电器产业株式会社 | Method and device for reducing block distortion |
| CN101770639A (en) * | 2010-01-14 | 2010-07-07 | 北京航空航天大学 | Enhancement method of low-illumination image |
Non-Patent Citations (2)
| Title |
|---|
| 吴铁洲 等: "直方图匹配图像增强技术的算法研究与实现", 《湖北工业大学学报》, vol. 20, no. 2, 30 April 2005 (2005-04-30), pages 59 - 61 * |
| 朱丽娜 等: "基于Wallis滤波的改进小波去噪方法研究", 《地理空间信息》, vol. 4, no. 6, 31 December 2006 (2006-12-31), pages 55 - 56 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105260997A (en) * | 2015-09-22 | 2016-01-20 | 北京好运到信息科技有限公司 | Method for automatically obtaining target image |
| CN106650794A (en) * | 2016-11-24 | 2017-05-10 | 北京理工大学 | Method and system for eliminating highlight of image affected by highlight reflection on object surface |
| CN106650794B (en) * | 2016-11-24 | 2019-10-01 | 北京理工大学 | A kind of image bloom removing method influenced by body surface high light reflectivity and system |
| CN107622481A (en) * | 2017-10-25 | 2018-01-23 | 沈阳东软医疗系统有限公司 | Reduce the method, apparatus and computer equipment of CT picture noises |
| US10783614B2 (en) | 2017-10-25 | 2020-09-22 | Beijing Neusoft Medical Equipment Co., Ltd. | Denoising CT image |
| CN107622481B (en) * | 2017-10-25 | 2022-09-30 | 东软医疗系统股份有限公司 | Method and device for reducing CT image noise and computer equipment |
| CN108830330A (en) * | 2018-06-22 | 2018-11-16 | 西安电子科技大学 | Classification of Multispectral Images method based on self-adaptive features fusion residual error net |
| CN108830330B (en) * | 2018-06-22 | 2021-11-02 | 西安电子科技大学 | Multispectral image classification method based on self-adaptive feature fusion residual error network |
| CN109816615A (en) * | 2019-03-06 | 2019-05-28 | 腾讯科技（深圳）有限公司 | Image processing method, device, equipment and storage medium |
| CN109816615B (en) * | 2019-03-06 | 2022-12-16 | 腾讯科技（深圳）有限公司 | Image restoration method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103049886B (en) | 2015-10-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103049886A (en) | Image texture repair method and system | |
| CN110148095B (en) | Underwater image enhancement method and enhancement device | |
| CN106710517B (en) | LED display bearing calibration | |
| CN103531175B (en) | Splicing display device screen intensity uniformity tuning method and system | |
| EP2296114B1 (en) | Edge extraction apparatus, surveying equipment, and program | |
| CN103150708B (en) | Based on the image Quick demisting optimization method of black channel | |
| CN101571950B (en) | Image restoring method based on isotropic diffusion and sparse representation | |
| CA2497212A1 (en) | Image fusion system and method | |
| CN105528797A (en) | Optical image color consistency self-adaption processing and quick mosaic method | |
| CN104574266B (en) | Morphing based on contour line | |
| CN102280097A (en) | Display brightness and chrominance uniformity regulating method | |
| CN105701780A (en) | Remote sensing image processing method and system thereof | |
| CN105046643A (en) | Image splicing method based on brightness adaptive registration | |
| CN103400343A (en) | Method for compensating uneven brightness of bottom view image under nighttime infrared | |
| CN104091307A (en) | Frog day image rapid restoration method based on feedback mean value filtering | |
| CN110472527A (en) | A kind of track line detecting method and system | |
| CN111462022B (en) | Underwater image sharpness enhancement method | |
| JP5448056B2 (en) | Distorted image correction apparatus and method | |
| CN111338051A (en) | Automatic focusing method and system based on TFT liquid crystal panel | |
| CN107424136B (en) | Method for processing Gaussian blur of picture under Mac | |
| CN104504666A (en) | Tone mapping method based on Laplacian pyramid | |
| CN103824250B (en) | image tone mapping method based on GPU | |
| CN105139345A (en) | Automatic searching method of high-quality non-standard Gamma curve | |
| CN116596768A (en) | Detail enhancement method based on multi-brightness image fusion | |
| CN115953341A (en) | Night rocket visible light image and infrared image regional fusion method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230807 Address after: 430000 No. 01, floor 11, office tower unit, Guanggu financial center construction project, No. 668, Gaoxin Avenue, Donghu New Technology Development Zone, Wuhan, Hubei Province Patentee after: Fangzhu (Wuhan) Technology Co.,Ltd. Patentee after: FOUNDER INTERNATIONAL Co.,Ltd. Address before: 19th Floor, Zhongxin Building, No. 52 North Fourth Ring West Road, Haidian District, Beijing, 100871 Patentee before: Founder International Co.,Ltd. (Beijing) Patentee before: FOUNDER INTERNATIONAL Co.,Ltd. |
|
| TR01 | Transfer of patent right |