CN103761728A - Method and device for lens shading correction - Google Patents

Method and device for lens shading correction Download PDF

Info

Publication number
CN103761728A
CN103761728A CN201310739631.0A CN201310739631A CN103761728A CN 103761728 A CN103761728 A CN 103761728A CN 201310739631 A CN201310739631 A CN 201310739631A CN 103761728 A CN103761728 A CN 103761728A
Authority
CN
China
Prior art keywords
brightness value
equation
correction equation
sampled point
value
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
Application number
CN201310739631.0A
Other languages
Chinese (zh)
Other versions
CN103761728B (en
Inventor
王栋
秦汝增
孔晓东
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Luo Sanjie
Original Assignee
Huawei Technologies Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Huawei Technologies Co Ltd filed Critical Huawei Technologies Co Ltd
Priority to CN201310739631.0A priority Critical patent/CN103761728B/en
Priority claimed from CN201310739631.0A external-priority patent/CN103761728B/en
Publication of CN103761728A publication Critical patent/CN103761728A/en
Application granted granted Critical
Publication of CN103761728B publication Critical patent/CN103761728B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Studio Devices (AREA)

Abstract

An embodiment of the invention provides a method and device for lens shading correction. A smooth curved surface equation is set as a correction equation; according to preset intervals, an imaging region is sampled, and original brightness values of sampling points are obtained; the errors between target brightness values obtained by adopting fitting of the correction equation and the original brightness values are calculated; the weighted value of each sampling point is set, an unknown parameter value of the correction equation is calculated when the quadratic sum of the errors is minimum; each pixel in the imaging region is corrected by using the correction equation, storage space only needs to be provided for the unknown parameter value of the correction equation, and the number of occupied register resources is reduced; besides, function values of the correction equation are continuously derivable in the whole imaging region, the correction equation conforms to the rule of overall brightness distribution of the imaging region, the brightness of the imaging region is corrected smoothly, and the problem that the brightness on the boundary is abnormal because a block processing mechanism is adopted is solved.

Description

Method and device that a kind of camera lens brightness decay is proofreaied and correct
Technical field
The present invention relates to optical imagery processing technology field, particularly relate to method and device that a kind of camera lens brightness decay is proofreaied and correct.
Background technology
Optical lens is the image device of commonly using now, and based on its optical physics characteristic, under uniform photoenvironment, the image that optical lens becomes is in the brightness difference of different photosensitive regions.The image that optical lens becomes, decays to the brightness of surrounding imaging gradually from the central point of the photosensitive region of image, and the some brightness far away of the central point of image middle distance photosensitive region is lower.In order to reduce the true brightness of the image that optical lens becomes, need to proofread and correct the brightness decay of image, the method for this correction is called camera lens brightness decay and proofreaies and correct (Les Shading Correction, LSC).
The mechanism of currently used traditional LSC method based on piecemeal processing, imaging region is divided into regular grid, the each corresponding reference mark coefficient in four summits of each grid, set up the linear mapping relation of each grid and four reference mark coefficients, utilize this linear mapping to be related to the brightness decay in calibration grid.
While adopting above-mentioned LSC method to carry out attenuation correction, each reference mark coefficient needs to take a storage space, if when imaging region is divided into the grid of n × n, needs n2 storage space storage reference mark coefficient, and shared register resources is many.
Summary of the invention
In view of this, method and device that the embodiment of the present invention provides a kind of camera lens brightness decay to proofread and correct, demarcate the unknown parameter in correction equation, utilizes correction equation to proofread and correct, and only needs the unknown parameter in storage correction equation, reduces shared register resources.
A kind of method that embodiment of the present invention first aspect provides camera lens brightness decay to proofread and correct, described method comprises:
Set a smooth surface equation as correction equation;
According to predetermined interval, imaging region is sampled, obtain the original brightness value of sampled point;
The target brightness value that the described correction equation matching of calculating employing obtains and the error amount of described original brightness value;
Weighted value is set to each described sampled point, error of calculation quadratic sum is the unknown parameters ' value of hour correction equation;
Utilize described correction equation to proofread and correct each pixel in described imaging region.
In the possible implementation of the first of embodiment of the present invention first aspect, described smooth surface equation includes but not limited to:
Any one in Quadratic Surface Equation, cubic surface equation, sine function equation, cosine function equation and Gaussian function equation.
In the possible implementation of the second of embodiment of the present invention first aspect, described in also comprise after obtaining the original brightness value of sampled point:
The original brightness value normalization of sampled point is obtained to normalization brightness value;
The target brightness value that matching of calculating employing correction equation obtains and the error amount of original brightness value are:
The target brightness value that the matching of calculating employing correction equation obtains and the error amount of normalization brightness value.
In the third possible implementation of embodiment of the present invention first aspect, described in also comprise after obtaining the original brightness value of sampled point:
When correction equation is inverse proportion equation, original brightness value is carried out to inverse proportion conversion and obtain inverse proportion brightness value;
The target brightness value that matching of calculating employing correction equation obtains and the error amount of original brightness value are:
The target brightness value that the matching of calculating employing correction equation obtains and the error amount of inverse proportion brightness value.
In conjunction with embodiment of the present invention first aspect, to the third possible implementation of first aspect, the target brightness value that the matching of described calculating employing correction equation obtains and the error amount of original brightness value are:
V n 1 = B nt x t 1 ^ - l n 1 ;
Wherein,
Figure BDA0000448693030000032
the error amount matrix being listed as for n capable 1, for the coordinate figure of the sampled point of the capable t of the n row compute matrix in correction equation,
Figure BDA0000448693030000034
for the matrix of unknown parameter in the capable 1 row correction equation of t,
Figure BDA0000448693030000035
for the sampled point original brightness value matrix that n capable 1 is listed as, the number that n is sampled point, the number that t is unknown parameter.
In conjunction with embodiment of the present invention first aspect to the third possible implementation of first aspect,, describedly weighted value is set to each sampled point, the error of calculation quadratic sum unknown parameters ' value of hour correction equation is:
Calculate time correction equation unknown number parameter value;
Wherein, P is the weight matrix of n sampled point, and V is the error amount matrix that n capable 1 is listed as.
The device that embodiment of the present invention second aspect provides a kind of camera lens brightness decay to proofread and correct, described device comprises:
Setup unit, for setting a smooth surface equation as correction equation;
Acquiring unit, for imaging region being sampled according to predetermined interval, obtains the original brightness value of sampled point;
The first computing unit, adopts target brightness value that correction equation matching obtains and the error amount of original brightness value for calculating;
The second computing unit, for weighted value being set to each sampled point, error of calculation quadratic sum is the unknown parameters ' value of hour correction equation;
Correcting unit, for utilizing correction equation to proofread and correct each pixel in imaging region.
In the possible implementation of the first of embodiment of the present invention second aspect, described device also comprises:
Normalization unit, obtains normalization brightness value for the original brightness value normalization to sampled point;
The first computing unit is:
The first computation subunit, adopts target brightness value that correction equation matching obtains and the error amount of normalization brightness value for calculating.
In the possible implementation of the second of embodiment of the present invention second aspect, described device also comprises:
Inverse proportion converter unit, for when correction equation is inverse proportion equation, carries out inverse proportion conversion to original brightness value and obtains inverse proportion brightness value;
Described the first computing unit is:
The second computation subunit, adopts target brightness value that correction equation matching obtains and the error amount of inverse proportion brightness value for calculating.
In conjunction with embodiment of the present invention second aspect to the possible implementation of the second of second aspect,
Described the second computing unit is specifically for calculating
Figure BDA0000448693030000041
time correction equation unknown number parameter;
Wherein, P is the weight matrix of n sampled point, and V is the error amount matrix that n capable 1 is listed as.
As shown in the above, the present invention has following beneficial effect:
The method and apparatus that the embodiment of the present invention provides a kind of camera lens brightness decay to proofread and correct, described method comprises: set a smooth surface equation as correction equation; According to predetermined interval, imaging region is sampled, obtain the original brightness value of sampled point; The target brightness value that the matching of calculating employing correction equation obtains and the error of original brightness value; Weighted value is set to each sampled point, error of calculation quadratic sum is the unknown parameters ' value of hour correction equation; Utilize correction equation to proofread and correct each pixel in imaging region, the present invention provides storage space only need to the unknown parameters ' value of correction equation, according to location parameter, can determine correction equation, unknown parameter generally, in 10, has reduced shared register resources greatly; And, correction equation is non-vanishing at other sample point derivatives except the central point of imaging region, that is to say, the rule that correction equation coincidence imaging region overall brightness distributes, brightness to imaging region is smoothly proofreaied and correct, there will not be owing to adopting piecemeal treatment mechanism, the problem that boundary brightness is abnormal.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, to the accompanying drawing of required use in embodiment or description of the Prior Art be briefly described below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skills, do not paying under the prerequisite of creative work, can also obtain according to these accompanying drawings other accompanying drawing.
Fig. 1 is that grid division schematic diagram is proofreaied and correct in the brightness decay of prior art camera lens;
Fig. 2 is embodiment of the method one process flow diagram that a kind of camera lens brightness decay of the present invention is proofreaied and correct;
Fig. 3 is uniform grid schematic diagram of the present invention;
Fig. 4 is the image before proofreading and correct;
Fig. 5 is the image adopting after camera lens brightness decay bearing calibration of the present invention is proofreaied and correct;
Fig. 6 is device embodiment tri-structural representations that a kind of camera lens brightness decay of the present invention is proofreaied and correct.
Embodiment
The embodiment of the invention discloses method and device that a kind of camera lens brightness decay is proofreaied and correct, demarcate the unknown parameter in correction equation, utilize correction equation to proofread and correct, only need the unknown parameter in storage correction equation, reduce shared register resources, realized the level and smooth correction to imaging region.
At present, LSC method used in the prior art is divided into regular grid by imaging region, and as shown in Figure 1, to be divided into 6 × 6 grid as example, the intersection point of every row dividing line and row dividing line is a reference mark, and the reference mark coefficient at each reference mark is:
E p = L 0 L p - - - ( 1 )
Wherein, L 0for the brightness of the central point of imaging region, L pbe that p reference mark is the original brightness before correcting.
As shown in Figure 1, to being positioned at a pixel p of selected net region, according to corresponding four the reference mark coefficient (F in four of this net region summits (1,2,3 and 4) 1, F 2, F 3and F 4), adopt bilinear interpolation to obtain the correction coefficient F of this pixel in this net region pxl:
F pxl = F 1 × ( GW - dx ) × ( GH - dy ) + F 2 × dx × ( GH - dy ) + F 3 × ( GW - dx × ) × dy + F 4 × dx × dy GW × GH
Wherein, F 1, F 2, F 3and F 4for the reference mark parameter of upper left, net region, pixel p place, upper right, lower-left and bottom right, dx is the horizontal range of pixel p apart from reference mark, upper left, dy is the vertical range of pixel p apart from reference mark, upper left, GW is the width of net region, pixel p place, and GH is the height in net region, pixel p place.Here it should be noted that, be positioned at the F of all pixels of same grid 1, F 2, F 3and F 4identical.
Brightness to this pixel in grid is proofreaied and correct:
L adj = L ori × F pxl - - - ( 2 )
Wherein, L adjfor this pixel in selected net region is proofreaied and correct rear object brightness, L orifor this pixel is proofreaied and correct front original brightness, F pxlthe correction coefficient that adopts bilinear interpolation to obtain for this pixel.
From above-mentioned example, when imaging region is divided into 6 × 6 grid, one has 36 reference mark coefficients, need 36 storage spaces, and, because the pixel in each grid is to proofread and correct according to the determined correction coefficient of reference mark coefficient on four of grid summits, the border of grid can not guarantee smooth continuous after correction, and the pixel of net boundary easily occurs becoming clear abnormal after proofreading and correct.
In order to solve the problems of the technologies described above, the embodiment of the present invention has been carried method and the device that a kind of camera lens brightness decay is proofreaied and correct, and below in conjunction with accompanying drawing, the embodiment of the present invention is specifically described.
Embodiment mono-
Fig. 2 is embodiment of the method one process flow diagram that a kind of camera lens brightness decay of the present invention is proofreaied and correct, and described method comprises:
Step 201: set a smooth surface equation as correction equation.
The Descartes's right-handed coordinate system that has the coordinate axis of horizontal X and numerical value Y both direction to form as true origin foundation take the key store of imaging region, set a smooth surface equation as correction equation, described correction equation is not 0 in the first order derivative value of other point except (0,0) some coordinate system in addition.
Described smooth surface equation comprises any one in Quadratic Surface Equation, cubic surface equation, sine function equation, cosine function equation and Gaussian function equation, here it should be noted that, smooth surface equation is not limited in above-mentioned several, can also be reciprocal equation etc., repeat no longer one by one here.
In the embodiment of the present invention, as an example of Quadratic Surface Equation example, be specifically described, method and Quadratic Surface Equation that other smooth surface equation is realized camera lens brightness decay correction as correction equation are similar, repeat no longer one by one here.
The embodiment of the present invention is set Quadratic Surface Equation (3) as correction equation:
F(x,y)=a×x×x+b×y×y+c×x×y+d×x+e×y+f (3)
Wherein, a, b, c, d, e, f is unknown parameter.
Step 202: according to predetermined interval, imaging region is carried out to uniform sampling, obtain the original brightness value of sampled point.
For the correction equation that comprises 6 unknown parameters (3), the minimum value that need to list 6 equations and form system of equations and obtain 6 unknown parameters, the unknown parameters ' value causing for fear of accidental sampling error departs from optimum solution, and imaging region is carried out to uniform sampling according to predetermined interval.During sampling, can set up the grid of the capable m row of m, select the intersection point of ranks as sampled point, obtain the original brightness value of sampled point, set up n equation, adopt equation number much larger than the overdetermined equation of unknown parameter number, unknown parameter to be solved, wherein n=m 2.
In order to guarantee to obtain the value of unknown parameter, while setting up the grid of the capable m row of m, the grid that minimum is 3 × 3, sets up 9 equation solutions.The value of m is larger, and solving of unknown parameter is more accurate.In the embodiment of the present invention, as shown in Figure 3, the grid take 11 × 11 is specifically described as example.Certainly, can also set up 9 × 9,13 × 13 or 15 × 15 grid, repeat no longer one by one here.While generally setting up grid, the value of m is odd number, and the coordinate that can guarantee like this central point of imaging region is (0,0).
When the grid that the capable m of m row are set up in employing carries out uniform sampling, not limiting each grid length and width must be in full accord, for example: in Fig. 3, for fear of the accidental error of grid edge brightness, choose 0.05 replacement 0, choose 0.95 replacement 1.
The grid of setting up the capable m row of m except above-mentioned employing carries out the method for uniform sampling, can also adopt additive method to carry out uniform sampling, requires the whole region of the basic uniform fold of sampled point.
Step 203: the target brightness value that the described correction equation matching of calculating employing obtains and the error amount of described original brightness value.
Adopt formula (4) error of calculation value:
V n 1 = B nt x t 1 ^ - l n 1 - - - ( 4 ) ;
Wherein,
Figure BDA0000448693030000072
the error amount matrix being listed as for n capable 1, the matrix being listed as for n capable 1, for the coordinate figure of the sampled point of the capable t of the n row compute matrix in correction equation,
Figure BDA0000448693030000074
for the matrix of unknown parameter in the capable 1 row correction equation of t, for the sampled point original brightness value matrix that n capable 1 is listed as, the number that n is sampled point, the number that t is unknown parameter.
Take Quadratic Surface Equation as correction equation, be example, in Quadratic Surface Equation, have 6 unknown parameters, matrix
Figure BDA0000448693030000082
(x, y) is the coordinate figure of sampled point.When setting up 11 × 11 grid, matrix B is 121 row, the matrix of 6 row.
Take Quadratic Surface Equation as correction equation, be example, correction equation has a, b, c, d, e, f, 6 parameters, matrix x t 1 ^ = a b c d e f ,
Figure BDA0000448693030000084
it is the matrix that 6 row 1 are listed as.
Figure BDA0000448693030000085
for the sampled point original brightness value matrix that n capable 1 is listed as, also can be right
Figure BDA0000448693030000086
be normalized, adopt the ratio of the original brightness value of each sampled point and the brightness value of imaging region central point to replace original brightness value, simplify and calculate.After normalization, the target brightness value that the matching of calculating employing correction equation obtains and the error amount of normalization brightness value.
Figure BDA0000448693030000087
the sampled point original brightness value matrix being listed as for n capable 1, when correction equation is inverse proportion equation, also can be right carry out inverse proportion conversion, adopt the inverse of the original brightness value of each sampled point to replace original brightness value, improve and proofread and correct result.After inverse proportion conversion, the target brightness value that the matching of calculating employing correction equation obtains and the error amount of inverse proportion brightness value
Here it should be noted that, in concrete computation process, according to actual conditions, can be right
Figure BDA0000448693030000089
both be normalized, and also carried out inverse proportion conversion, the order that normalization and inverse proportion change does not limit.
Step 204: weighted value is set to each described sampled point, error of calculation quadratic sum is the unknown parameters ' value of hour described correction equation.
According to sampled point, the significance level of proofreading and correct result is arranged to different weight P, weight can be divided into multiple ranks, for near sampled value corner and central point, adopt large weight, for other regions, adopt little weight, weight can specifically be set according to the reality situation of proofreading and correct.
Adopt the unknown parameter of hour correction equation of formula (5) error of calculation quadratic sum:
∂ PV 2 ∂ x ^ = ∂ V T PV ∂ x ^ = 0 - - - ( 5 )
By formula (5) formula (6) of deriving to obtain:
V TPB=0 (6)
Bring formula (4) into formula (6) and obtain formula (7):
B T PB x ^ - B T Pl = 0 - - - ( 7 )
Matrix of unknown parameters
Figure BDA0000448693030000093
can calculate according to formula (8):
x ^ = ( B T PB ) - 1 B T Pl - - - ( 8 )
Utilize formula (8) to calculate the value of the unknown parameter in correction equation.
Step 205: utilize described correction equation to proofread and correct each pixel in described imaging region.
Determine after the location parameter in correction equation, obtain correction equation F(x, y), according to formula (9), each pixel in imaging region is proofreaied and correct:
L adj ( x , y ) = Lori ( x , y ) F ( x , y ) - - - ( 9 )
Wherein, F(x, y) be correction coefficient, L adj(x, y) is the target brightness value after proofread and correct at (x, y) some place, and Lori (x, y) is the original brightness value before proofread and correct at (x, y) some place.
As shown in the above, the present invention has following beneficial effect:
Set a smooth surface equation as correction equation; According to predetermined interval, imaging region is carried out to uniform sampling, obtain the original brightness value of sampled point; The target brightness value that the matching of calculating employing correction equation obtains and the error of original brightness value; Weighted value is set to each sampled point, error of calculation quadratic sum is the unknown parameter of hour correction equation; Utilize correction equation to proofread and correct each pixel in imaging region, the present invention provides storage space only need to the unknown parameter of correction equation, according to location parameter, can determine correction equation, unknown parameter generally, in 10, has reduced shared register resources greatly; And correction equation functional value can be led at whole imaging region everywhere continuous, that is to say, the rule that correction equation coincidence imaging region overall brightness distributes, brightness to imaging region is smoothly proofreaied and correct, and there will not be owing to adopting piecemeal treatment mechanism, the problem that boundary brightness is abnormal.
Embodiment bis-
A kind of camera lens brightness decay of the present invention bearing calibration scene embodiment, in embodiment bis-, adopts Quadratic Surface Equation as correction equation, sets up 11 × 11 grid and realizes uniform sampling, and described method comprises:
(1) set Quadratic Surface Equation as correction equation;
Correction equation is F (x, y)=a × x × x+b × y × y+c × x × y+d × x+e × y+f.
(2) set up 11 × 11 grid and realize uniform sampling, as shown in Figure 3
For fear of the impact of the accidental error at imaging region edge, center position is got 0.05 replacement 0; Marginal point is got 0.95 replacement 1.
Table 1 sampling point position coordinate
X -0.9 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 0.9
Y -0.9 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 0.9
Table 2 sampled point original brightness value
073 079 093 107 114 114 112 103 093 082 073
076 080 097 111 118 113 113 108 098 082 077
084 089 106 124 133 132 129 122 109 090 083
087 089 111 135 140 138 134 130 115 094 086
092 102 120 144 161 165 158 143 123 102 092
093 098 121 144 159 157 153 142 124 101 091
092 098 120 143 158 158 153 140 121 098 093
088 090 112 133 144 137 135 130 116 094 087
081 085 105 122 128 125 120 119 108 089 081
075 081 097 109 119 116 117 111 098 084 075
071 077 092 106 110 111 109 102 092 080 072
The corresponding sampled point of each brightness value is identical with the grid intersection point position of 11 × 11 shown in Fig. 3, for example, brightness value 073 correspondence in form 2 upper left corners be in Fig. 3, to be positioned at (0.9, 0.9) brightness value of the sampled point of locating, upper right corner brightness value 073 correspondence be in Fig. 3, to be positioned at (0.9, 0.9) brightness value of the sampled point of locating, lower left corner brightness value 071 correspondence be in Fig. 3, to be positioned at (0.9,-0.9) brightness value of the sampled point of locating, lower right corner brightness value 072 correspondence be in Fig. 3, to be positioned at (0.9,-0.9) brightness value of the sampled point of locating, the coordinate of the sampled point that other brightness value is corresponding can be determined according to the method described above, here repeat no longer one by one.
In actual computation process, the sampled point original brightness value normalization in can his-and-hers watches 2, obtains normalization brightness value matrix l.Can also be when correction equation be inverse proportion equation, the original brightness of the sampled point in his-and-hers watches 2 carries out inverse proportion conversion and obtains inverse proportion brightness value matrix l.Adopt normalization and inverse proportion conversion, can, when solving the unknown parameter of correction equation, reduce calculated amount, improve computational accuracy.
(3) target brightness value that the matching of calculating employing correction equation obtains and the error amount of original brightness value, build the equation that solves unknown parameter in correction equation.
x ^ = ( B T PB ) - 1 B T Pl
Wherein, matrix P is weight matrix, can on the size that affects of image rectification, to sampled point, weighted value be set according to each sampled point.Provide the instantiation that weighted value is set below, the weight of 4 ranks is set, WPN represents standard weight 1, and WPL represents that weight is that 10, WXL represents that weight is that 100, XXL represents that weight is 100.Here it should be noted that, for the setting of the weight of sampled point, be not limited in the above-mentioned form of expression, the concrete numerical value of more weight rank and weight can also be set according to actual conditions, repeat no longer one by one here.
Table 3 graticule mesh sampled point weighted value
XXL WPL WPN WPN WPN WPL WPN WPN WPN WPL XXL
WPL WXL WPN WPN WPN WPN WPN WPN WPN WXL WPL
WPN WPN WPN WPN WPN WPN WPN WPN WPN WPN WPN
WPN WPN WPN WPN WPN WPN WPN WPN WPN WPN WPN
WPN WPN WPN WPN WPL WPN WPL WPN WPN WPN WPN
WPL WPN WPN WPN WPN XXL WPN WPN WPN WPN WPL
WPN WPN WPN WPN WPL WPN WPL WPN WPN WPN WPN
WPN WPN WPN WPN WPN WPN WPN WPN WPN WPN WPN
WPN WPN WPN WPN WPN WPN WPN WPN WPN WPN WPN
WPL WXL WPN WPN WPN WPN WPN WPN WPN WXL WPL
XXL WPL WPN WPN WPN WPL WPN WPN WPN WPL XXL
The corresponding sampled point of each weighted value is identical with the grid intersection point position of 11 × 11 shown in Fig. 3, for example, what the weighted value XXL in form 2 upper left corners was corresponding is in Fig. 3, to be positioned at (0.9,0.9) weighted value of the sampled point of locating, what upper right corner weighted value XXL was corresponding is in Fig. 3, to be positioned at (0.9,0.9) weighted value of the sampled point of locating, the coordinate of the sampled point that other brightness value is corresponding can be determined according to the method described above, repeats no longer one by one here.
Be given in below in 11 × 11 grid, each sampled point according to
Figure BDA0000448693030000122
the matrix B of calculating gained, one has 121 groups of data:
Table 4 graticule mesh sampled point quadratic polynomial parameter value
Figure BDA0000448693030000121
Figure BDA0000448693030000131
Figure BDA0000448693030000141
Figure BDA0000448693030000151
Figure BDA0000448693030000171
In table 4, the numeral before every line parameter in bracket obtains the coordinate of the sampled point of this line parameter value.
(4) coefficient of the correction equation obtaining is according to the method described above:
Table 5 correction equation unknown parameters ' value
Unknown parameter a b c d e f
Value 1.505827 0.854009 -0.014956 -0.017219 0.035240 1.616404
(5) unknown parameter of the correction equation of confirming according to (4) obtains correction equation, and reason utilizes correction equation to proofread and correct each pixel in imaging region.Fig. 4 is the imaging region before proofreading and correct, and Fig. 5 is the imaging region after proofreading and correct, and can find out that the brightness branch of Fig. 5 is more even than the brightness part of Fig. 4.
Embodiment tri-
Fig. 6 is device embodiment tri-structural representations that a kind of camera lens brightness decay of the present invention is proofreaied and correct, and is and the corresponding device of method described in embodiment mono-that described device comprises:
Setup unit 601, for setting a smooth surface equation as correction equation.
Acquiring unit 602, for imaging region being carried out to uniform sampling according to predetermined interval, obtains the original brightness value of sampled point.
The first computing unit 603, adopts target brightness value that correction equation matching obtains and the error amount of original brightness value for calculating.
The second computing unit 604, for weighted value being set to each sampled point, error of calculation quadratic sum is the unknown parameter of hour correction equation.
In concrete implementation, described the second computing unit 604 is specifically for calculating
Figure BDA0000448693030000181
time correction equation unknown number parameter;
Wherein, P is the weight matrix of n sampled point, and V is the error amount matrix that n capable 1 is listed as.
Correcting unit 605, for utilizing correction equation to proofread and correct each pixel in imaging region.
In concrete implementation procedure, described device also comprises:
Normalization unit, obtains normalization brightness value for the original brightness value normalization to sampled point;
The first computing unit 603 is:
The first computation subunit, adopts target brightness value that correction equation matching obtains and the error amount of normalization brightness value for calculating.
In concrete implementation procedure, described device also comprises:
Inverse proportion converter unit, for when correction equation is inverse proportion equation, carries out inverse proportion conversion to original brightness value and obtains inverse proportion brightness value;
Described the first computing unit 603 is:
The second computation subunit, adopts target brightness value that correction equation matching obtains and the error amount of inverse proportion brightness value for calculating.
Similar with embodiment mono-herein, the description of reference example one, repeats no more here.
Embodiment tetra-
For the device embodiment tetra-that a kind of camera lens brightness decay of the present invention is proofreaied and correct, described device comprises first memory and first processor, and described first memory is used for storing instruction, and described first processor is used for transferring instruction, and described instruction comprises:
Set a smooth curve equation as correction equation;
According to predetermined interval, imaging region is carried out to uniform sampling, obtain the original brightness value of sampled point;
The target brightness value that the described correction equation matching of calculating employing obtains and the error amount of described original brightness value;
Weighted value is set to each described sampled point, error of calculation quadratic sum is the unknown parameters ' value of positive equation described in hour school;
Utilize described correction equation to proofread and correct each pixel in described imaging region.
Optionally, described smooth curve equation comprises:
Any one in quadratic curve equation, cubic curve equation, sine function equation, cosine function equation and Gaussian function equation.
Optionally, after obtaining the original brightness value of sampled point described in, also comprise:
The original brightness value normalization of sampled point is obtained to normalization brightness value;
The target brightness value that matching of calculating employing correction equation obtains and the error amount of original brightness value are:
The target brightness value that the matching of calculating employing correction equation obtains and the error amount of normalization brightness value.
Optionally, after obtaining the original brightness value of sampled point described in, also comprise:
When correction equation is inverse proportion equation, original brightness value is carried out to inverse proportion conversion and obtain inverse proportion brightness value;
The target brightness value that matching of calculating employing correction equation obtains and the error amount of original brightness value are:
The target brightness value that the matching of calculating employing correction equation obtains and the error amount of inverse proportion brightness value.
Optionally, the target brightness value that the matching of described calculating employing correction equation obtains and the error amount of original brightness value are:
V n 1 = B nt x t 1 ^ - l n 1 ;
Wherein,
Figure BDA0000448693030000192
the error amount matrix being listed as for n capable 1, the matrix being listed as for n capable 1,
Figure BDA0000448693030000193
for the coordinate figure of the sampled point of the capable t of the n row compute matrix in correction equation,
Figure BDA0000448693030000194
for the matrix of unknown parameter in the capable 1 row correction equation of t,
Figure BDA0000448693030000195
for the sampled point original brightness value matrix that n capable 1 is listed as, the number that n is sampled point, the number that t is unknown parameter.
Optionally, describedly weighted value is set to each sampled point, the error of calculation quadratic sum unknown parameters ' value of hour correction equation is:
Calculate
Figure BDA0000448693030000201
time correction equation unknown number parameter value;
Wherein, P is the weight matrix of n sampled point, and V is the error amount matrix that n capable 1 is listed as.
The foregoing is only preferred embodiment of the present invention, in order to limit the present invention, within the spirit and principles in the present invention not all, any modification of doing, be equal to replacement, improvement etc., within all should being included in protection scope of the present invention.

Claims (10)

1. the method that camera lens brightness decay is proofreaied and correct, is characterized in that, described method comprises:
Set a smooth surface equation as correction equation;
According to predetermined interval, imaging region is sampled, obtain the original brightness value of sampled point;
The target brightness value that the described correction equation matching of calculating employing obtains and the error amount of described original brightness value;
Weighted value is set to each described sampled point, error of calculation quadratic sum is the unknown parameter of hour correction equation;
Utilize described correction equation to proofread and correct each pixel in described imaging region.
2. method according to claim 1, is characterized in that, described smooth surface equation includes but not limited to:
Any one in Quadratic Surface Equation, cubic surface equation, sine function equation, cosine function equation and Gaussian function equation.
3. method according to claim 1, is characterized in that, described in also comprise after obtaining the original brightness value of sampled point:
The original brightness value normalization of sampled point is obtained to normalization brightness value;
The target brightness value that matching of calculating employing correction equation obtains and the error amount of original brightness value are:
The target brightness value that the matching of calculating employing correction equation obtains and the error amount of normalization brightness value.
4. method according to claim 1, is characterized in that, described in also comprise after obtaining the original brightness value of sampled point:
When correction equation is inverse proportion equation, original brightness value is carried out to inverse proportion conversion and obtain inverse proportion brightness value;
The target brightness value that matching of calculating employing correction equation obtains and the error amount of original brightness value are:
The target brightness value that the matching of calculating employing correction equation obtains and the error amount of inverse proportion brightness value.
5. according to the method described in claim 1-4 any one, it is characterized in that, the target brightness value that the matching of described calculating employing correction equation obtains and the error amount of original brightness value are:
V n 1 = B nt x t 1 ^ - l n 1 ;
Wherein, the error amount matrix being listed as for n capable 1,
Figure FDA0000448693020000023
for the coordinate figure of the sampled point of the capable t of the n row compute matrix in correction equation, for the matrix of unknown parameter in the capable 1 row correction equation of t,
Figure FDA0000448693020000025
for the sampled point original brightness value matrix that n capable 1 is listed as, the number that n is sampled point, the number that t is unknown parameter.
6. according to the method described in claim 1-4 any one, it is characterized in that, describedly weighted value is set to each sampled point, the error of calculation quadratic sum unknown parameters ' value of hour correction equation is:
Calculate
Figure FDA0000448693020000026
time correction equation unknown number parameter value;
Wherein, P is the weight matrix of n sampled point, and V is the error amount matrix that n capable 1 is listed as.
7. the device that camera lens brightness decay is proofreaied and correct, is characterized in that, described device comprises:
Setup unit, for setting a smooth surface equation as correction equation;
Acquiring unit, for imaging region being sampled according to predetermined interval, obtains the original brightness value of sampled point;
The first computing unit, adopts target brightness value that correction equation matching obtains and the error amount of original brightness value for calculating;
The second computing unit, for weighted value being set to each sampled point, error of calculation quadratic sum is the unknown parameters ' value of hour correction equation;
Correcting unit, for utilizing correction equation to proofread and correct each pixel in imaging region.
8. device according to claim 7, is characterized in that, described device also comprises:
Normalization unit, obtains normalization brightness value for the original brightness value normalization to sampled point;
The first computing unit is:
The first computation subunit, adopts target brightness value that correction equation matching obtains and the error amount of normalization brightness value for calculating.
9. device according to claim 7, is characterized in that, described device also comprises:
Inverse proportion converter unit, for when correction equation is inverse proportion equation, carries out inverse proportion conversion to original brightness value and obtains inverse proportion brightness value;
Described the first computing unit is:
The second computation subunit, adopts target brightness value that correction equation matching obtains and the error amount of inverse proportion brightness value for calculating.
10. according to the device described in claim 7-9 any one, it is characterized in that,
Described the second computing unit is specifically for calculating
Figure FDA0000448693020000031
time correction equation unknown number parameter;
Wherein, P is the weight matrix of n sampled point, and V is the error amount matrix that n capable 1 is listed as.
CN201310739631.0A 2013-12-27 A kind of method and device of camera lens brightness decay correction Active CN103761728B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310739631.0A CN103761728B (en) 2013-12-27 A kind of method and device of camera lens brightness decay correction

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310739631.0A CN103761728B (en) 2013-12-27 A kind of method and device of camera lens brightness decay correction

Publications (2)

Publication Number Publication Date
CN103761728A true CN103761728A (en) 2014-04-30
CN103761728B CN103761728B (en) 2016-11-30

Family

ID=

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110084856A (en) * 2019-04-24 2019-08-02 Oppo广东移动通信有限公司 Luminance regulating method, device, electronic equipment and the storage medium of uncalibrated image
CN113592712A (en) * 2021-07-30 2021-11-02 北京百度网讯科技有限公司 Image processing method, device, equipment, storage medium and cloud VR system
CN114078426A (en) * 2021-09-30 2022-02-22 卡莱特云科技股份有限公司 Method for eliminating influence of camera curve on lamp point image and display screen correction method
CN115100071A (en) * 2022-07-18 2022-09-23 芯原微电子(上海)股份有限公司 Brightness balance correction method and device, image acquisition equipment and storage medium
CN116777999A (en) * 2023-06-28 2023-09-19 深圳市度申科技有限公司 Multi-adaptability high-level flat field correction method for area array camera
CN116777999B (en) * 2023-06-28 2024-07-26 深圳市度申科技有限公司 Multi-adaptability high-level flat field correction method for area array camera

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2007075066A1 (en) * 2005-12-29 2007-07-05 Mtekvision Co., Ltd. Image processor, apparatus and method for lens shading compensation
CN101166235A (en) * 2006-10-16 2008-04-23 索尼株式会社 Lens apparatus, image capture apparatus, and method for correcting image quality
KR20110072213A (en) * 2009-12-22 2011-06-29 엘지이노텍 주식회사 Image correction method
CN102131040A (en) * 2010-06-04 2011-07-20 苹果公司 Adaptive lens shading correction

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2007075066A1 (en) * 2005-12-29 2007-07-05 Mtekvision Co., Ltd. Image processor, apparatus and method for lens shading compensation
CN101166235A (en) * 2006-10-16 2008-04-23 索尼株式会社 Lens apparatus, image capture apparatus, and method for correcting image quality
KR20110072213A (en) * 2009-12-22 2011-06-29 엘지이노텍 주식회사 Image correction method
CN102131040A (en) * 2010-06-04 2011-07-20 苹果公司 Adaptive lens shading correction

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110084856A (en) * 2019-04-24 2019-08-02 Oppo广东移动通信有限公司 Luminance regulating method, device, electronic equipment and the storage medium of uncalibrated image
CN110084856B (en) * 2019-04-24 2021-07-27 Oppo广东移动通信有限公司 Method and device for adjusting brightness of calibration image, electronic equipment and storage medium
CN113592712A (en) * 2021-07-30 2021-11-02 北京百度网讯科技有限公司 Image processing method, device, equipment, storage medium and cloud VR system
CN114078426A (en) * 2021-09-30 2022-02-22 卡莱特云科技股份有限公司 Method for eliminating influence of camera curve on lamp point image and display screen correction method
CN115100071A (en) * 2022-07-18 2022-09-23 芯原微电子(上海)股份有限公司 Brightness balance correction method and device, image acquisition equipment and storage medium
CN116777999A (en) * 2023-06-28 2023-09-19 深圳市度申科技有限公司 Multi-adaptability high-level flat field correction method for area array camera
CN116777999B (en) * 2023-06-28 2024-07-26 深圳市度申科技有限公司 Multi-adaptability high-level flat field correction method for area array camera

Similar Documents

Publication Publication Date Title
CN100405398C (en) Method for obtaining image by vertically correcting satellite
CN103473771B (en) A kind of camera scaling method
CN109816732B (en) Calibration method, calibration system, correction method, correction system and vehicle
CN106780321B (en) CBERS-02 satellite HR sensor image overall tight orientation and correction splicing method
CN106570907B (en) Camera calibration method and device
CN103150724B (en) Segmented model-based camera calibration method
CN102661717A (en) Monocular vision measuring method for iron tower
CN112862895B (en) Fisheye camera calibration method, device and system
CN104463791A (en) Fisheye image correction method based on spherical model
CN104820984B (en) A kind of satellite remote sensing three line scanner stereopsis processing system and method
CN104200454A (en) Fisheye image distortion correction method and device
CN105118086A (en) 3D point cloud data registering method and system in 3D-AOI device
CN105069761A (en) Low-calculation cylindrical surface panoramagram realization method and system
CN102968788A (en) Wave band registering method based on regular grid surface element
CN103247033A (en) Vector data correction method for image-based control point selection
CN103116878A (en) Method and device for correcting image barrel distortion and image processing device
CN116086411B (en) Digital topography generation method, device, equipment and readable storage medium
CN109544642A (en) A kind of TDI-CCD camera parameter scaling method based on N-type target
CN104180794B (en) The disposal route in digital orthoimage garland region
CN104103057A (en) Method for fast generating panorama by fish eye image
CN103124334A (en) Lens distortion correction method
CN110298778B (en) Traffic cell demographic method based on raster data and area ratio correction
CN105571598B (en) A kind of assay method of laser satellite altimeter footmark camera posture
CN103778610A (en) Geometric pretreatment method for vertical rail swing images of satellite-borne linear array sensor
CN107945136B (en) Fisheye image correction method, fisheye image correction system, fisheye image correction equipment and computer storage medium

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
TR01 Transfer of patent right

Effective date of registration: 20201224

Address after: 518000 Baoan District Xin'an street, Shenzhen, Guangdong, No. 625, No. 625, Nuo platinum Plaza,

Patentee after: SHENZHEN SHANGGE INTELLECTUAL PROPERTY SERVICE Co.,Ltd.

Address before: 518129 Bantian HUAWEI headquarters office building, Longgang District, Guangdong, Shenzhen

Patentee before: HUAWEI TECHNOLOGIES Co.,Ltd.

Effective date of registration: 20201224

Address after: No.6, block B, Taoyuan Road, Shuyang Industrial Park, Suqian City, Jiangsu Province

Patentee after: Luo Sanjie

Address before: 518000 Baoan District Xin'an street, Shenzhen, Guangdong, No. 625, No. 625, Nuo platinum Plaza,

Patentee before: SHENZHEN SHANGGE INTELLECTUAL PROPERTY SERVICE Co.,Ltd.