CN102957916B - Color information interpolating method - Google Patents
Color information interpolating method Download PDFInfo
- Publication number
- CN102957916B CN102957916B CN201110243976.8A CN201110243976A CN102957916B CN 102957916 B CN102957916 B CN 102957916B CN 201110243976 A CN201110243976 A CN 201110243976A CN 102957916 B CN102957916 B CN 102957916B
- Authority
- CN
- China
- Prior art keywords
- variability
- color information
- pixel
- vertical
- level
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Landscapes
- Facsimile Image Signal Circuits (AREA)
- Image Processing (AREA)
- Color Image Communication Systems (AREA)
Abstract
The invention discloses a color information interpolating method. The color information interpolating method includes: receiving luminance information of a pixel matrix, wherein the luminance information is in Bayer pattern array and records specific color information of each pixel of the pixel matrix, and the specific color information is one of red information, green information and blue information; computing a horizontal detail variation amount and a vertical detail variation amount of each pixel according to the luminance information; and interpolating each pixel in the luminance information by one of a horizontal color information estimation value, a vertical color information estimation value and a non-directional color information estimation value according to the horizontal detail variation amount and the vertical detail variation amount, so that color information in each pixel except for the specific color information is obtained.
Description
Technical field
The present invention relates to a kind of color information interpolating method, espespecially one can utilize image contrast to adjust and judge, to increase the color information interpolating method of image detail walking direction with details amount of variability and aberration variation average magnitude.
Background technology
In general, for saving cost, the picture element matrix of consumer image sensor is when sensing image is to produce luminance information, this luminance information is with Bel figure (Bayer pattern) arrangement, namely only note down a specific color information of each pixel in picture element matrix, this specific color information is one in the middle of a red color information, a green tint information and a blue color information.Therefore, when an image processor receives this luminance information, need to utilize interpolation (Interpolation) computing to carry out the color information re-establishing lost, make each pixel all have three kinds of color informations.
In known technology, mainly in the middle of fixing use bilinearity interpolating method (bilinear), border interpolating method and minimum colour difference assessment interpolating method, one carries out interpolation.With bilinearity interpolating method, it is that the color information of loss is utilized Bilinear Method interpolation, and main shortcoming is to cause image fog and serious false colour phenomenon, and in edge (edge) easily misjudgment; With border interpolating method, it is the judgement travel direction interpolation color information of loss being utilized level and vertical direction, main shortcoming is not high to the direction determining accuracy on trickle border, such as cannot know that directivity can misjudgment at image high frequency treatment, and occur on image clathrate (as originally for be laterally judged as directly to); And with minimum colour difference assessment interpolating method, it the color information lost is utilized to inject line direction interpolation with the minimum look of neighborhood pixels, main shortcoming is not high to image high frequency treatment direction determining accuracy.
From the above, using one in the middle of bilinearity interpolating method, border interpolating method and minimum colour difference assessment interpolating method because known technology is fixing when carrying out interpolation, therefore can cause image fog or judging that accuracy is not high at image high frequency treatment.In view of this, real necessity having improvement of known technology.
Summary of the invention
Therefore, namely main purpose of the present invention is to provide one that image contrast can be utilized to adjust and judges, to increase the color information interpolating method of image detail walking direction with details amount of variability and aberration variation average magnitude.
The present invention discloses a kind of color information interpolating method, include the luminance information receiving and correspond to a picture element matrix, this luminance information schemes arrangement with a Bel, this luminance information notes down a specific color information of each pixel in this picture element matrix, and this specific color information is one in the middle of a red color information, a green tint information and a blue color information; According to this luminance information, calculate a level detail amount of variability and a vertical detail amount of variability of this pixel; And according to this level detail amount of variability and this vertical detail amount of variability, interpolation is carried out to this pixel in this luminance information, to obtain the color information except this specific color information with one in the middle of a level color information estimated value, a vertical color information estimated value and a directionless color information estimated value.
Coordinate following schemes, the detailed description of embodiment and claims at this, by address after other object of the present invention and advantage be specified in.
Accompanying drawing explanation
1A figure is the schematic diagram of the embodiment of the present invention one image processor.
1B figure is when in 1A figure, picture element matrix is 5*5, the schematic diagram of a luminance information.
Fig. 2 is the schematic diagram that in Fig. 1, an image contrast adjusting device is used for carrying out a function of image contrast adjustment.
Fig. 3 A and Fig. 3 B is the schematic diagram of the embodiment of the present invention one color information interpolation flow process.
Fig. 4 is the schematic diagram of a detail calculation unit calculated level details amount of variability and vertical detail amount of variability in Fig. 1.
Fig. 5 is the schematic diagram of an aberration computing unit calculated level aberration variation average magnitude and vertical aberration variation average magnitude in Fig. 1.
Fig. 6 is the schematic diagram of another color information interpolation flow process of the embodiment of the present invention.
Wherein, description of reference numerals is as follows:
10 image processors
12 image sensors
102 image contrast adjusting devices
104 detail calculation unit
106 Colorimetry unit
108 interpolation judging units
30,60 flow processs
300 ~ 330,600 ~ 608 steps
LI luminance information
CLI contrast luminance information
HDV, HDV
1~ HDV
3level detail amount of variability
VDV, VDV
1~ VDV
3vertical detail amount of variability
CHavg horizontal aberration variation average magnitude
CVavg vertical aberration variation average magnitude
R, G, B pixel
F (x) function
R11 ~ R55, G12 ~ G54, B22 ~ B44 pixel
Embodiment
Please refer to 1A figure, 1A figure is the schematic diagram of the embodiment of the present invention one image processor 10.Image processor 10 is mainly used in, as image capture units such as digital camera, network camera (internet protocol camera), external video camera or computer built-in cameras, including image contrast adjusting device 102, detail calculation unit 104, aberration computing unit 106 and an interpolation judging unit 108.In simple terms, image processor 10 can receive a luminance information LI corresponding to a picture element matrix PM by an image sensor 12.Luminance information LI is with Bel figure (Bayer pattern) arrangement, a specific color information PCI of each pixel in its recording pixel matrix PM.Specific color information PCI is one in the middle of a red color information, a green tint information and a blue color information; For example, 1B figure is picture element matrix PM when being 5*5, the schematic diagram of luminance information LI, wherein pixel R represents that its specific color information PCI is red color information, pixel G represents that its specific color information PCI is green tint information, and pixel B represents that its specific color information PCI is blue color information.Detail calculation unit 104 can according to luminance information LI, calculate a level detail amount of variability HDV and a vertical detail amount of variability VDV of a pixel PX, and interpolation judging unit 108 can according to level detail amount of variability HDV, vertical detail amount of variability VDV, interpolation is carried out to pixel PX in luminance information LI, to obtain the color information in pixel PX except specific color information PCI with one in the middle of a level color information estimated value Eh, an a vertical color information estimated value Ev and directionless color information estimated value En.
In the case, the level detail amount of variability HDV of pixel PX deduct vertical detail amount of variability VDV be greater than a details threshold values DT (threshold) time, interpolation judging unit 108 can carry out interpolation to pixel PX in luminance information LI by its vertical color information estimated value Ev, and the specific color information PCI namely when judging that pixel PX horizontal direction changes greatly with its vertical direction surrounding pixel carries out interpolation; The vertical detail amount of variability VDV of pixel PX deduct level detail amount of variability HDV be greater than details threshold values DT time, interpolation judging unit 108 can carry out interpolation to pixel PX in luminance information LI by its level color information estimated value Eh, and the specific color information PCI namely when judging that pixel PX vertical direction changes greatly with its horizontal direction surrounding pixel carries out interpolation.
On the other hand, the level detail amount of variability HDV of pixel PX deduct vertical detail amount of variability VDV and vertical detail amount of variability VDV deduct level detail amount of variability HDV be all less than details threshold values DT time, Colorimetry unit 106 can according to luminance information LI, one horizontal aberration variation average magnitude CHavg and the vertical aberration variation average magnitude CVavg of calculating pixel PX, make interpolation judging unit 108 according to horizontal aberration variation average magnitude CHavg and vertical aberration variation average magnitude CVavg, with level color information estimated value Eh, in the middle of vertical color information estimated value Ev and directionless color information estimated value En, one carries out interpolation to pixel PX in luminance information LI, to obtain the color information in pixel PX except specific color information PCI.
Specifically, the horizontal aberration variation average magnitude CHavg of pixel PX deduct vertical aberration variation average magnitude CVavg be greater than difference limen value CT of the same colour time, interpolation judging unit 108 can carry out interpolation to pixel PX in luminance information LI by its vertical color information estimated value Ev, and the specific color information PCI namely when judging that pixel PX horizontal direction changes greatly with its vertical direction surrounding pixel carries out interpolation; The vertical aberration variation average magnitude CVavg of pixel PX deduct horizontal aberration variation average magnitude CHavg be greater than details threshold values CT time, interpolation judging unit 108 can carry out interpolation to pixel PX in luminance information LI by its level color information estimated value Eh, and the specific color information PCI namely when judging that pixel PX vertical direction changes greatly with its horizontal direction surrounding pixel carries out interpolation; And deduct vertical aberration variation average magnitude CVavg and vertical aberration variation average magnitude CVavg at the horizontal aberration variation average magnitude CHavg of pixel PX and deduct (ascending the throne in image flat region) when horizontal aberration variation average magnitude CHavg is less than details threshold values CT, interpolation judging unit 108 can carry out interpolation to pixel PX in luminance information LI by its directionless color information estimated value En, namely carries out interpolation when judging that pixel PX changes greatly without specific direction with the specific color information PCI of its surrounding pixel.Thus, image processor 10 can judge according to the variation of details amount of variability HDV, VDV and aberration average magnitude CHavg, CVavg, then carries out interpolation with suitable color information estimated value, to increase image detail walking direction.
In addition, please refer to Fig. 2, Fig. 2 is the schematic diagram that in Fig. 1, image contrast adjusting device 102 is used for carrying out a function F (x) of image contrast adjustment.As shown in Figure 2, image processor 10 can also comprise image contrast adjusting device 102, be used for after image processor 10 receives luminance information LI, according to function F (x), image contrast adjustment is carried out to luminance information LI, to promote the contrast of low brightness pixel in picture element matrix PM, and then generation one contrast luminance information CLI gives detail calculation unit 104 and Colorimetry unit 106, make its respectively according to this make a variation average magnitude CHavg and vertical aberration of calculated level details amount of variability HDV and vertical detail amount of variability VDV and horizontal aberration to make a variation average magnitude CVavg, to increase details walking direction.Notably, interpolation judging unit 108 is still the pixel of script luminance information LI but not the pixel of contrast luminance information CLI carries out interpolation, and to rebuild color information, therefore image pixel can keep the pure of raw information and edge details can not be affected.Thus, the contrast of low brightness pixel during image contrast adjusting device 102 can promote, with low-light level image detail walking direction in increasing.
Specifically, in one embodiment, the operation of image processor 10 can be a color information interpolation flow process 30, and as shown in Fig. 3 A and Fig. 3 B, it comprises the following steps:
Step 300: start.
Step 302: according to pixel PX in luminance information LI and all pixels of surrounding thereof, a level detail amount of variability HDV of calculating pixel PX
1and a vertical detail amount of variability VDV
1.
Step 304: determined level details amount of variability HDV
1deduct vertical detail amount of variability VDV
1whether be greater than a details threshold values DT
1.If so, carry out step 326; If not, carry out step 306.
Step 306: judge vertical detail amount of variability VDV
1deduct level detail amount of variability HDV
1whether be greater than details threshold values DT
1.If so, carry out step 328; If not, carry out step 308.
Step 308: according to pixel PX in luminance information LI and around and there is for interpolation color information the pixel of same hue information, a level detail amount of variability HDV of calculating pixel PX
2and a vertical detail amount of variability VDV
2.
Step 310: determined level details amount of variability HDV
2deduct vertical detail amount of variability VDV
2whether be greater than a details threshold values DT
2.If 326, carry out step; If not, carry out step 312.
Step 312: judge vertical detail amount of variability VDV
2deduct level detail amount of variability HDV
2whether be greater than details threshold values DT
2.If so, carry out step 328; If not, carry out step 314.
Step 314: according to pixel PX in luminance information LI and around and there is for interpolation color information the pixel of different color information, a level detail amount of variability HDV of calculating pixel PX
3and a vertical detail amount of variability VDV
3.
Step 316: determined level details amount of variability HDV
3deduct vertical detail amount of variability VDV
3whether be greater than a details threshold values DT
3.If so, carry out step 326; If not, carry out step 318.
Step 318: judge vertical detail amount of variability VDV
3deduct level detail amount of variability HDV
3whether be greater than details threshold values DT
3.If so, carry out step 328; If not, carry out step 320.
Step 320: according to pixel PX in luminance information LI and vertical and horizontal direction pixel thereof, the horizontal aberration variation average magnitude CHavg of calculating pixel PX and vertical aberration variation average magnitude CVavg.
Step 322: determined level aberration variation average magnitude CHavg deducts vertical aberration variation average magnitude CVavg and whether is greater than aberration threshold values CT.If so, step is carried out; If not 326, carry out step 318.
Step 324: judge that vertical aberration variation average magnitude CVavg deducts horizontal aberration variation average magnitude CHavg and whether is greater than aberration threshold values CT.If so, step is carried out; If not 328, carry out step 330.
Step 326: interpolation is carried out to pixel PX in luminance information LI with vertical color information estimated value Ev.
Step 328: interpolation is carried out to pixel PX in luminance information LI with level color information estimated value Eh.
Step 330: interpolation is carried out to pixel PX in luminance information LI with its directionless color information estimated value En.
From color information interpolation flow process 30, the present invention is in determined level amount of variability or vertically amount of variability whichever is obviously larger, during to determine to use vertical color information estimated value Ev, level color information estimated value Eh or directionless color information estimated value En to carry out interpolation, first by roughly to trickle, three stage computational details amount of variability HDV
1~ HDV
3, VDV
1~ VDV
3, with determined level amount of variability or vertical amount of variability whichever obviously comparatively large, judge with aberration variation average magnitude CHavg, CVavg again if all cannot judge.
For example, please refer to Fig. 4, Fig. 4 is detail calculation unit 104 calculated level details amount of variability HDV in Fig. 1
1~ HDV
3and vertical detail amount of variability VDV
1~ VDV
3schematic diagram.According to color information interpolation flow process 30, when for reconstructing green tint information G33 that a pixel R33 that a specific color information PCI33 is red color information loses, understand first according to pixel R33 in luminance information LI and all pixels of surrounding (as ground-color portion) thereof, the level detail amount of variability HDV of calculating pixel R33
1and vertical detail amount of variability VDV
1, as follows:
HDV
1＝k1*|G32-G34|+k2*|α*R33-R31-R35|+(k3*|B22-B24|+k4*|β*G23-G21-G25|)/k5+(k6*|B42-B44|+k7*|γ*G43-G41-G45|)/k8
VDV
1＝k1*|G23-G43|+k2*|α*R33-R13-R53|+(k3*|B22-B42|+k4*|β*G32-G12-G52|)/k5+(k6*|B24-B44|+k7*|γ*G34-G14-G54|)/k8
Wherein, k1, k2, k3, k4, k5, k6, k7, k8 and α, β, γ are default parameter, HDV
1be the details amount of variability of first horizontal direction, VDV
1it is the details amount of variability of first vertical direction.
From above formula, level detail amount of variability HDV
1centered by pixel R33, the amount of variability of calculating pixel G32 and G34 and pixel R33, R31 and R35, and to close from the locus of center pixel R33 according to its horizontal level be give different weights, adjacent horizontal pixel G21, B22, G23, B24, G25 and G41, B42, G43, B44, G45 all adopt this account form, finally the level detail amount of variability of diverse location are added up and can obtain level detail amount of variability HDV
1.The rest may be inferred, the vertical detail amount of variability of diverse location can be added up to obtain vertical detail amount of variability VDV
1.
Then, if determined level details amount of variability HDV
1deduct vertical detail amount of variability VDV
1be greater than details threshold values DT
1, then with vertical color information estimated value Ev, interpolation is carried out to pixel R33 in luminance information LI, to rebuild green tint information G33; If judge vertical detail amount of variability VDV
1deduct level detail amount of variability HDV
1be greater than details threshold values DT
1, then with level color information estimated value Eh, interpolation is carried out to pixel R33 in luminance information LI, to rebuild green tint information G33.Finally, if judge vertical detail amount of variability VDV
1deduct level detail amount of variability HDV
1and level detail amount of variability HDV
1deduct vertical detail amount of variability VDV
1all be less than details threshold values DT
1, according to pixel R33 in luminance information LI and around and there is for interpolation color information G33 the pixel of same hue information (green), the level detail amount of variability HDV of calculating pixel R33
2and vertical detail amount of variability VDV
2, as follows:
HDV
2＝g1*|G12-G14|+g2*|G21-G23|+g3*|G23-G25|+g4*|G32-G34|+g5*|G41-G43|+|G45-G43|+g6*|G52-G54|
VDV
2＝g1*|G21-G41|+g2*|G12-G32|+g3*|G32-G52|+g4*|G23-G43|+g5*|G14-G34|+|G54-G34|+g6*|G25-G45
Wherein, g1, g2, g3, g4, g5, g6 are default parameter, HDV
2be the details amount of variability of second horizontal direction, VDV
2it is the details amount of variability of second water Vertical dimension.
From above formula, level detail amount of variability HDV
2centered by pixel R33, calculate the amount of variability of adjacent level green pixel, as pixel G32 and G34, pixel G21 and G23, pixel G23 and G25 etc., and give different weights according to its horizontal level from the spatial relation of center pixel R33, adjacent horizontal pixel all adopts this account form, finally the level detail amount of variability of diverse location is added up and can obtain level detail amount of variability HDV
2.The rest may be inferred, the vertical detail amount of variability of diverse location can be added up to obtain vertical detail amount of variability VDV
2.
Then, if determined level details amount of variability HDV
2deduct vertical detail amount of variability VDV
2be greater than details threshold values DT
2, then with vertical color information estimated value Ev, interpolation is carried out to pixel R33 in luminance information LI, to rebuild green tint information G33; If judge vertical detail amount of variability VDV
1deduct level detail amount of variability HDV
2be greater than details threshold values DT
2, then with level color information estimated value Eh, interpolation is carried out to pixel R33 in luminance information LI, to rebuild green tint information G33.Finally, if judge vertical detail amount of variability VDV
2deduct level detail amount of variability HDV
2and level detail amount of variability HDV
2deduct vertical detail amount of variability VDV
2all be less than details threshold values DT
2, according to pixel R33 in luminance information LI and around and there is for interpolation color information G33 the pixel of different color information (red and blue), the level detail amount of variability HDV of calculating pixel R33
3and vertical detail amount of variability VDV
3, as follows:
HDV
3＝r1*|B22-B24|+r2*|B42-B44|+r3*|R31-R33|+r4*|R35-R33|
VDV
3＝r1*|B22-B42|+r2*|B24-B44|+r3*|R13-R33|+r4*|R53-R33|
Wherein, r1, r2, r3, r4 are default parameter, HDV
3be the details amount of variability of the 3rd horizontal direction, VDV
3it is the details amount of variability of the 3rd vertical direction.
From above formula, level detail amount of variability HDV
3centered by pixel R33, calculate the amount of variability of adjacent level blue pixel and red pixel, as pixel R31 and R33, pixel R33 and R35, pixel B 22 and B24 etc., and give different weights according to its horizontal level from the spatial relation of center pixel R33, adjacent horizontal pixel all adopts this account form, finally the level detail amount of variability of diverse location is added up and can obtain level detail amount of variability HDV
3.The rest may be inferred, the vertical detail amount of variability of diverse location can be added up to obtain vertical detail amount of variability VDV
3.
Then, if determined level details amount of variability HDV
3deduct vertical detail amount of variability VDV
3be greater than details threshold values DT
3, then with vertical color information estimated value Ev, interpolation is carried out to pixel R33 in luminance information LI, to rebuild green tint information G33; If judge vertical detail amount of variability VDV
3deduct level detail amount of variability HDV
3be greater than details threshold values DT
3, then with level color information estimated value Eh, interpolation is carried out to pixel R33 in luminance information LI, to rebuild green tint information G33.
Finally, please refer to Fig. 5, Fig. 5 is the schematic diagram of Colorimetry unit 106 calculated level aberration variation average magnitude CHavg and vertical aberration variation average magnitude CVavg in Fig. 1.If judge vertical detail amount of variability VDV
3deduct level detail amount of variability HDV
3and level detail amount of variability HDV
3deduct vertical detail amount of variability VDV
3all be less than details threshold values DT
3, according to pixel R33 in luminance information LI and vertical and horizontal direction pixel (as ground-color portion), the horizontal aberration variation average magnitude CHavg of calculating pixel R33 and vertical aberration variation average magnitude CVavg, as follows:
CH1＝|G32-R33|
CH2＝|G32-R31|
CH3＝|G34-R33|
CH4＝|G34-R35|
CV1＝|G23-R33|
CV2＝|G23-R13|
CV3＝|G43-R33|
CV4＝|G43-R53|
CHavg＝(v1*CH1+v2*CH2+v3*CH3+v4*CH4)/v5
CVavg＝(v1*CV1+v2*CV2+v3*CV3+v4*CV4)/v5
Wherein, CH1, CH2, CH3, CH4 are neighbor other horizontal direction aberration amount of variability, CV1, CV2, CV3, CV4 are neighbor other vertical direction aberration amount of variability, CHavg is the aberration variation average magnitude of horizontal direction, CVavg is the aberration variation average magnitude of vertical direction, and v1, v2, v3, v4, v5 are default parameter.
From above formula, horizontal aberration amount of variability is centered by pixel R33, the aberration amount of variability of calculating pixel G32 and R33, pixel G32 and R31, pixel G34 and R33, pixel G34 and R35, and give different weights according to its horizontal level from the spatial relation of center pixel R33, adjacent horizontal pixel all adopts this account form, finally the horizontal aberration amount of variability of diverse location is added up and average to obtain horizontal aberration variation average magnitude CHavg.The rest may be inferred, the vertical detail amount of variability of diverse location can be added up to obtain vertical aberration variation average magnitude CVavg.
Then, if determined level aberration variation average magnitude CHavg deducts vertical aberration variation average magnitude CVavg be greater than aberration threshold values CT, then with vertical color information estimated value Ev, interpolation is carried out to pixel R33 in luminance information LI, to rebuild green tint information G33; If judge, vertical aberration variation average magnitude CVavg deducts horizontal aberration variation average magnitude CHavg and is greater than aberration threshold values CT, then carry out interpolation with level color information estimated value Eh to pixel R33 in luminance information LI, to rebuild green tint information G33; If horizontal aberration variation average magnitude CHavg deducts vertical aberration variation average magnitude Cvavg and vertical aberration variation average magnitude CVavg and deducts horizontal aberration variation average magnitude CHavg and be less than aberration threshold values CT, then judge that pixel R33 position is in a flat region, with directionless color information estimated value En, interpolation is carried out to pixel R33 in luminance information LI, to rebuild green tint information G33.
Wherein, level color information estimated value Eh be in the horizontal direction pixel of pixel R33 in luminance information LI with have for interpolation color information the pixel (green) of same hue information a level average and have different color information pixel (red and blue) a horizontal amount of variability with, vertical color information estimated value Ev be in the vertical direction pixel of pixel R33 in luminance information LI with have for interpolation color information the pixel of same hue information a vertical average and have different color information pixel a vertical amount of variability with, and directionless color information En be in the horizontal direction of pixel R33 in luminance information LI and vertical direction pixel with a directionless mean value for interpolation color information with the pixel of same hue information, as follows:
Eh＝(G32+G34)/2+(p1*R33-R31-R35)/p2
Ev＝(G23+G43)/2+(p1*R33-R13-R53)/p2
En＝(G32+G34+G23+G43)/4
Wherein, p1, p2 are default parameter, and Eh is the color information estimated value of horizontal direction, and Ev is the color information estimated value of vertical direction, and En is nondirectional color information estimated value.
From above formula, level color information estimated value Eh is the summation of amount of variability of the mean value of calculating pixel G32 and G34 and pixel R33, R31, R35, vertical color information estimated value Ev is the summation of amount of variability of the mean value of calculating pixel G23 and G43 and pixel R33, R13, R53, and nondirectional color information estimated value En is the mean value of calculating pixel G32, G34, G23, G43.
It should be noted that main spirits of the present invention is to judge according to level and vertical detail amount of variability and the aberration average magnitude that makes a variation, then suitably color information estimated value carries out interpolation, to increase image detail walking direction.Those of ordinary skill in the art when modifying according to this or change, and are not limited thereto.For example, in above-described embodiment, image processor 10 comprises image contrast adjusting device 102, with the contrast of low brightness pixel in lifting, and then be convenient to judge, image processor 10 also can not comprise image contrast adjusting device 102 in actual applications, and only detail calculation unit 104 and Colorimetry unit 106 carry out calculating but not contrast luminance information CLI according to luminance information LI; In addition, in embodiment all cannot with details amount of variability determined level amount of variability or vertical amount of variability whichever obviously larger, judge with aberration variation average magnitude, also can not judge with aberration variation average magnitude in actual applications, directly carry out interpolation with nondirectional color information estimated value En; Moreover, with by roughly to trickle in color information interpolation flow process 30, three stage computational details amount of variability HDV
1~ HDV
3, VDV
1~ VDV
3, with determined level amount of variability or vertical amount of variability whichever obviously comparatively large, in fact also only can calculate central one and judge to the two, and account form also be not limited to above-mentioned formula, as long as concept is carried out according to this.
Therefore, the color information interpolation operation of image processor 10, can be summarized as a color information interpolation flow process 60, as shown in Figure 6, it comprises the following steps:
Step 600: start.
Step 602: receive the luminance information LI corresponding to picture element matrix PM, luminance information LI schemes arrangement with Bel, luminance information LI notes down the specific color information PCI of each pixel in picture element matrix PM, and specific color information PCI is one in the middle of a red color information, a green tint information and a blue color information.
Step 604: according to luminance information LI, calculates level detail amount of variability HDV and the vertical detail amount of variability VDV of a pixel PX.
Step 606: according to level detail amount of variability HDV and vertical detail amount of variability VDV, interpolation is carried out, to obtain the color information in pixel PX except specific color information PCI with pixel PX in one luminance information LI in the middle of a level color information estimated value Eh, an a vertical color information estimated value Ev and directionless color information estimated value En.
Step 608: terminate.
The detailed content of color information interpolation flow process 60 with reference to the relevant running of above-mentioned image processor 10, can not repeat them here.
In known technology, owing to being one in the middle of fixing use bilinearity interpolating method, border interpolating method and minimum colour difference assessment interpolating method when carrying out interpolation, therefore can causing image fog or judging that accuracy is not high at image high frequency treatment.In comparison, the present invention judges according to level and vertical detail amount of variability and the aberration average magnitude that makes a variation, and then suitably color information estimated value carries out interpolation, to increase image detail walking direction.In addition, the present invention more first can carry out image contrast adjustment, and with the contrast of low brightness pixel in promoting, then calculated level and vertical detail amount of variability and aberration make a variation average magnitude, low-light level image detail walking direction in therefore can increasing.
The foregoing is only the preferred embodiments of the present invention, all equalizations done according to the claims in the present invention change and modify, and all should belong to covering scope of the present invention.
Claims (8)
1. a color information interpolating method, is characterized in that, includes:
Receive the luminance information corresponding to a picture element matrix, this luminance information schemes arrangement with a Bel, this luminance information notes down a specific color information of each pixel in this picture element matrix, and this specific color information is one in the middle of a red color information, a green tint information and a blue color information;
According to this luminance information, calculate a level detail amount of variability and a vertical detail amount of variability of a pixel; And
According to this level detail amount of variability and this vertical detail amount of variability, interpolation is carried out to this pixel in this luminance information, to obtain the color information in this pixel except this specific color information with one in the middle of a level color information estimated value, a vertical color information estimated value and a directionless color information estimated value;
Wherein, according to this luminance information, the step of this level detail amount of variability and this vertical detail amount of variability of calculating this pixel includes:
According to this pixel in this luminance information and all pixels of surrounding thereof, calculate one first level detail amount of variability and the one first vertical detail amount of variability of this pixel; And
This first level detail amount of variability deduct this first vertical detail amount of variability or this first vertical detail amount of variability deduct this first level detail amount of variability be all less than one first details threshold values time, according to this pixel in this luminance information and around and there is for interpolation color information the pixel of same hue information, one second level detail amount of variability and the one second vertical detail amount of variability of this pixel is calculated.
2. color information interpolating method as claimed in claim 1, it is characterized in that, according to this level detail amount of variability and this vertical detail amount of variability, with one in the middle of this level color information estimated value, this vertical color information estimated value and this directionless color information estimated value, the step that this pixel in this luminance information carries out interpolation is included:
This level detail amount of variability deduct this vertical detail amount of variability or this vertical detail amount of variability deduct this level detail amount of variability be greater than a details threshold values time, with this vertical color information estimated value or this level color information estimated value, interpolation is carried out to this pixel in this luminance information respectively.
3. color information interpolating method as claimed in claim 1, it is characterized in that, according to this level detail amount of variability and this vertical detail amount of variability, with one in the middle of this level color information estimated value, this vertical color information estimated value and this directionless color information estimated value, the step that this pixel in this luminance information carries out interpolation is included:
This level detail amount of variability deduct this vertical detail amount of variability or this vertical detail amount of variability deduct this level detail amount of variability be all less than a details threshold values time, according to this pixel in this luminance information and vertical and horizontal direction pixel thereof, calculate a horizontal aberration variation average magnitude and a vertical aberration variation average magnitude of this pixel; And
According to this horizontal aberration variation average magnitude and this vertical aberration variation average magnitude, interpolation is carried out to this pixel in this luminance information, to obtain the color information in this pixel except this specific color information with one in the middle of this level color information estimated value, this vertical color information estimated value and this directionless color information estimated value.
4. color information interpolating method as claimed in claim 1, it is characterized in that, according to this luminance information, the step of this level detail amount of variability and this vertical detail amount of variability of calculating this pixel includes:
Image contrast adjustment is carried out to this luminance information, to produce a contrast luminance information; And
According to this contrast luminance information, calculate this level detail amount of variability and this vertical detail amount of variability of this pixel.
5. color information interpolating method as claimed in claim 1, it is characterized in that, according to this level detail amount of variability and this vertical detail amount of variability, with one in the middle of this level color information estimated value, this vertical color information estimated value and this directionless color information estimated value, the step that this pixel in this luminance information carries out interpolation is included:
This first level detail amount of variability deduct this first vertical detail amount of variability or this first vertical detail amount of variability deduct this first level detail amount of variability be greater than one first details threshold values time, with this vertical color information estimated value or this level color information estimated value, interpolation is carried out to this pixel in this luminance information respectively.
6. color information interpolating method as claimed in claim 1, it is characterized in that, according to this level detail amount of variability and this vertical detail amount of variability, with one in the middle of this level color information estimated value, this vertical color information estimated value and this directionless color information estimated value, the step that this pixel in this luminance information carries out interpolation is included:
This second level detail amount of variability deduct this second vertical detail amount of variability or this second vertical detail amount of variability deduct this second level detail amount of variability be greater than one second details threshold values time, with this vertical color information estimated value or this level color information estimated value, interpolation is carried out to this pixel in this luminance information respectively.
7. color information interpolating method as claimed in claim 1, it is characterized in that, according to this luminance information, the step of this level detail amount of variability and this vertical detail amount of variability of calculating this pixel includes:
This second level detail amount of variability deduct this second vertical detail amount of variability or this second vertical detail amount of variability deduct this second level detail amount of variability be all less than one first details threshold values time, according to this pixel in this luminance information and around and there is for interpolation color information the pixel of different color information, one the 3rd level detail amount of variability and one the 3rd vertical detail amount of variability of this pixel is calculated.
8. color information interpolating method as claimed in claim 1, it is characterized in that, this level color information estimated value be in the horizontal direction pixel of this pixel in this luminance information with have for interpolation color information the pixel of same hue information a level average and have different color information pixel a horizontal amount of variability and, this vertical color information estimated value be in the vertical direction pixel of this pixel in this luminance information with have for interpolation color information the pixel of same hue information a vertical average and have different color information pixel a vertical amount of variability with, and this directionless color information estimated value be in the horizontal direction of this pixel in this luminance information and vertical direction pixel with a directionless mean value for interpolation color information with the pixel of same hue information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110243976.8A CN102957916B (en) | 2011-08-24 | 2011-08-24 | Color information interpolating method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110243976.8A CN102957916B (en) | 2011-08-24 | 2011-08-24 | Color information interpolating method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102957916A CN102957916A (en) | 2013-03-06 |
| CN102957916B true CN102957916B (en) | 2015-01-07 |
Family
ID=47766072
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110243976.8A Expired - Fee Related CN102957916B (en) | 2011-08-24 | 2011-08-24 | Color information interpolating method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102957916B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1722852A (en) * | 2004-03-15 | 2006-01-18 | 微软公司 | High-Quality Gradient-Corrected Linear Interpolation for Color Image Demosaicing |
-
2011
- 2011-08-24 CN CN201110243976.8A patent/CN102957916B/en not_active Expired - Fee Related
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1722852A (en) * | 2004-03-15 | 2006-01-18 | 微软公司 | High-Quality Gradient-Corrected Linear Interpolation for Color Image Demosaicing |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102957916A (en) | 2013-03-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6381215B2 (en) | Image processing apparatus, image processing method, display apparatus, display apparatus control method, and program | |
| CN102665030B (en) | Improved bilinear Bayer format color interpolation method | |
| US8441538B2 (en) | Image generation apparatus and image generation method | |
| US10984504B2 (en) | Advanced demosaicing with angle compensation and defective pixel correction | |
| CN107851307A (en) | To the method and system of the Bayer types of image data demosaicing for image procossing | |
| US20090180687A1 (en) | Method for realizing color interpolation of value correction of exceptional points | |
| US8749673B2 (en) | Image generation device and image generation system, method and program | |
| JP6172935B2 (en) | Image processing apparatus, image processing method, and image processing program | |
| CN108009997B (en) | Method and device for adjusting image contrast | |
| CN103079028A (en) | Black edge detection method and image processing device of image screen | |
| JP2018517993A (en) | Real-time video enhancement method, terminal, and non-transitory computer-readable storage medium | |
| JP4253655B2 (en) | Color interpolation method for digital camera | |
| CN105141838A (en) | Demosaicing method and device using demosaicing method | |
| CN105430357B (en) | The demosaicing methods and device of imaging sensor | |
| CN103841388A (en) | Mosaic removing method and device | |
| CN102957916B (en) | Color information interpolating method | |
| CN109285113B (en) | Improved color image interpolation method based on gradient | |
| TWI464704B (en) | Color information interpolation method | |
| TW201445506A (en) | Method and apparatus for determining interpolating direction for color demosaicking and non-transitory machine-readable storage medium | |
| CN103313066B (en) | Interpolation method and device | |
| JP2017126964A (en) | Imaging control unit, and imaging apparatus | |
| JP2011155365A (en) | Image processing apparatus and image processing method | |
| CN104079901A (en) | Image processing apparatus and method, and program | |
| CN103037268B (en) | Method and device for watermark adding | |
| JP2010028374A (en) | Image processor, method of interpolating image signal, and image processing program |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20150107 Termination date: 20170824 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |