CN113658054A - Infrared image splicing correction method based on temperature drift characteristic line approximation - Google Patents

Infrared image splicing correction method based on temperature drift characteristic line approximation Download PDF

Info

Publication number
CN113658054A
CN113658054A CN202110763276.5A CN202110763276A CN113658054A CN 113658054 A CN113658054 A CN 113658054A CN 202110763276 A CN202110763276 A CN 202110763276A CN 113658054 A CN113658054 A CN 113658054A
Authority
CN
China
Prior art keywords
temperature drift
infrared image
matrix
row
method based
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
CN202110763276.5A
Other languages
Chinese (zh)
Other versions
CN113658054B (en
Inventor
王小勇
文高进
吴立民
钟灿
龙亮
王哲
黄浦
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.)
Beijing Institute of Space Research Mechanical and Electricity
Original Assignee
Beijing Institute of Space Research Mechanical and Electricity
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 Beijing Institute of Space Research Mechanical and Electricity filed Critical Beijing Institute of Space Research Mechanical and Electricity
Priority to CN202110763276.5A priority Critical patent/CN113658054B/en
Publication of CN113658054A publication Critical patent/CN113658054A/en
Application granted granted Critical
Publication of CN113658054B publication Critical patent/CN113658054B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G06T5/80—
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/20—Image enhancement or restoration by the use of local operators
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/50—Image enhancement or restoration by the use of more than one image, e.g. averaging, subtraction
    • G06T5/70—
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10048—Infrared image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20212—Image combination
    • G06T2207/20221—Image fusion; Image merging

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Radiation Pyrometers (AREA)

Abstract

The invention relates to an infrared image splicing correction method based on temperature drift characteristic line approximation, belonging to the field of infrared image processing; the invention realizes the simultaneous temperature drift characteristic calculation and non-uniformity correction processing in the infrared image splicing processing, overcomes the defect of the traditional infrared image splicing method, provides a multi-scale approximation method of the relative temperature drift characteristic curve of the infrared image, forms a more stable and more effective infrared image splicing method, can effectively remove the temperature drift noise in the infrared image, furthest keeps the high-frequency information of the image and obtains a high-quality infrared splicing result image.

Description

Infrared image splicing correction method based on temperature drift characteristic line approximation
Technical Field
The invention belongs to the field of infrared image processing, and relates to an infrared image splicing correction method based on temperature drift characteristic line approximation.
Background
On one hand, the response transfer function of each pixel of the infrared detector is different due to the influence of the manufacturing process and materials, so that stripe noise, called non-uniformity noise, exists in the obtained infrared image. On the other hand, under the influence of environmental temperature, target heat radiation and the like, the temperature of the infrared detector can change continuously, temperature drift phenomena of different degrees are formed, the spectral responsivity of the detector can be changed, the output voltage of the detector is changed, and therefore the repeatability and the accuracy of a measuring result are influenced. When a plurality of detectors work cooperatively, the formed images have serious inconsistency, which brings difficulty to the splicing of infrared images. The image quality is seriously reduced by the two factors, and the wide application of infrared imaging in the fields of medical treatment, monitoring, agriculture and forestry and the like is greatly hindered, so that when the combined infrared remote sensing of a plurality of detectors is carried out, the temperature drift-based processing splicing correction is required to be carried out to improve the quality of the infrared image.
The infrared image splicing method mainly comprises two methods, namely a radiation thermometer calibration method and an inter-chip heterogeneity correction method. The radiation thermometer calibration method adopts the radiation thermometer to measure the temperature of the detector in real time to obtain a temperature drift curve of spectral responsivity, obtains a fitting equation of the temperature drift curve through data fitting, determines a functional relation between the temperature of the detector and the spectral responsivity and a temperature drift correction coefficient, and realizes correction of the spectral responsivity temperature drift of the detector. The method needs to add a radiation thermometer to each infrared detector, which brings difficulties to the structural design and the reliability design of the detection system, and has larger limitation in practical application. The inter-slice nonuniformity correction method generally performs correction processing by using a two-point or multi-point nonuniformity correction method. The two methods can not effectively solve the problem of non-linear temperature drift of the multiple detectors, and the problem of chromatic aberration still exists between corrected images.
Disclosure of Invention
The technical problem solved by the invention is as follows: the method overcomes the defects of the prior art, provides the infrared image splicing correction method based on the temperature drift characteristic line approximation, can effectively remove the temperature drift noise in the infrared image, furthest keeps the high-frequency information of the image, and obtains the high-quality infrared splicing result image.
The technical scheme of the invention is as follows:
an infrared image splicing correction method based on temperature drift characteristic line approximation comprises the following steps:
step one, setting an initialization scale parameter d and a correction error e of a first iteration1；
Inputting the infrared image X of m rows and n columns and the infrared image Y of s rows and n columns, and calculating the extended image of the infrared image X
Figure BDA0003150762790000021
Calculating an extended image of the infrared image Y
Figure BDA0003150762790000022
Step three, calculating a multi-scale filter H and a temperature drift initial curve V;
step four, according to the extended image
Figure BDA0003150762790000023
Expanding an image
Figure BDA0003150762790000024
Carrying out approximation correction optimization processing on the multi-scale filter H and the temperature drift initial curve V to obtain a final correction error ed；
Step five, pair edMaking a judgment when ed＜ed-1The temperature drift curve of the d iteration
Figure BDA0003150762790000025
Record as the temperature drift optimization curve
Figure BDA0003150762790000026
And recording the value of dTo an optimum scale
Figure BDA0003150762790000027
Entering the step six; when e isd≥ed-1Directly entering the step six;
step six, setting a threshold value D, comparing the value D with the threshold value D, and entering step seven when D is larger than D; otherwise, returning to the step two;
step seven, optimizing a curve e according to the temperature driftd＜ed-1Optimal size of the optical fiber
Figure BDA00031507627900000215
Carrying out approximate splicing processing on the infrared image X and the infrared image Y to obtain a spliced result graph
Figure BDA0003150762790000028
In the above infrared image stitching correction method based on temperature drift characteristic line approximation, in the first step, d is 2, e1＝1。
In the above infrared image stitching correction method based on temperature drift characteristic line approximation, in the second step,
Figure BDA0003150762790000029
is a matrix with m rows and n +2d-2 columns;
Figure BDA00031507627900000210
is a matrix with s rows and n +2d-2 columns; wherein the image is expanded
Figure BDA00031507627900000211
Satisfies the following iterative formula:
Figure BDA00031507627900000212
wherein j is 1,2, … …, m; t-1, 2, … …, d-1;
Figure BDA00031507627900000213
wherein j is 1,2, …, m; t-1, 2, … …, d-1;
Figure BDA00031507627900000214
wherein j is 1,2, … …, m; t is 1,2, … …, n;
expanding an image
Figure BDA0003150762790000031
Satisfies the following iterative formula:
Figure BDA0003150762790000032
wherein j is 1,2, … …, s; t-1, 2, … …, d-1;
Figure BDA0003150762790000033
wherein j is 1,2, … …, s; t-1, 2, … …, d-1;
Figure BDA0003150762790000034
wherein j is 1,2, … …, s; t is 1,2, … …, n.
In the above infrared image stitching correction method based on temperature drift characteristic line approximation, in the third step, the multi-scale filter H is a 2d-1 dimensional vector, specifically:
H＝[12……d*(d-1)……1]/(d+d*(d-1))
the initial curve V of the temperature drift is:
V＝[V1 V2…Vn+2d-2]
in the formula, Xk,pThe value of the k row and the p column in the matrix X;
Yk,pthe values of the k row and the p column in the matrix Y are shown;
Figure BDA0003150762790000037
is a matrix
Figure BDA0003150762790000038
The value of the kth row and the value of the pth column;
Figure BDA0003150762790000039
is a matrix
Figure BDA00031507627900000310
The value of the kth row and the value of the pth column;
Hkis the kth element in vector H;
Vkis the kth element in vector V;
Figure BDA0003150762790000035
in the above infrared image stitching correction method based on temperature drift characteristic line approximation, in the fourth step, a specific approach to the correction optimization processing is as follows:
s41, setting the iteration number as a, and setting the correction error r of the 0 th iteration0Is 1;
s42, updating and calculating a temperature drift filtering parameter matrix U of the row a and the column n +2 d-2;
s43, calculating an approximation graph Z of S +1 rows and n columns and a corrected error r of the a-th iterationa；
S44, correcting error r for a (a) th iterationaIs judged when r isa＜ra-1In time, the temperature drift optimization curve under the current scale parameter d
Figure BDA0003150762790000036
Updating the element to the a-th row of the temperature drift filtering parameter matrix U, and adding raIs expressed as the final correction error edGo to S45; when r isa≥ra-1When the update is not performed, the flow proceeds directly to S45;
s45, p | ra-ra-1| making a judgment when | ra-ra-1|＜1.0×10-5Then, entering the step five; otherwise, let a be a +1, return is made to S42.
In the above infrared image stitching correction method based on temperature drift characteristic line approximation, in S42, the calculation method of the temperature drift filtering parameter matrix U is:
when a is 1, initializing, making U in 1 st row and j th column of matrix U1,j＝Vj，j＝1，2，……，n+2d-2；
Updating U1,jThe calculation formula of (2) is as follows:
Figure BDA0003150762790000041
wherein j is d, d +1, … …, d + n-1;
in the formula of Uk,pThe values of the k row and the p column in the matrix U are shown;
Hkis the kth element in vector H;
Vkis the kth element in vector V;
when a is more than 1, initializing the Ua,j＝Ua-1,j，j＝1，2，……，n+2d-2；
Updating Ua,jThe calculation formula of (2) is as follows:
Figure BDA0003150762790000042
in the above infrared image stitching correction method based on temperature drift characteristic line approximation, in S43, the approximation graph Z satisfies:
Z1,j＝Xm,jwherein j is 1,2, … …, n
Figure BDA0003150762790000043
Wherein t is 1,2, … …, s; j is 1,2, … …, n;
corrected error r for iteration aaComprises the following steps:
Figure BDA0003150762790000044
in the formula, Zk,pIs the k-th row in the matrix Z,The value in column p.
In the above infrared image stitching correction method based on temperature drift characteristic line approximation, in the sixth step, the value range of the threshold value D is as follows: d is more than or equal to 12 and less than or equal to 16.
In the above infrared image stitching correction method based on temperature drift characteristic line approximation, in the seventh step, the stitched result graph
Figure BDA0003150762790000045
Satisfies the following conditions:
Figure BDA0003150762790000046
wherein i is 1,2, … …, m; j is 1,2, … …, n;
Figure BDA0003150762790000047
wherein i is 1,2, … …, s; j is 1,2, … …, n;
in the formula (I), the compound is shown in the specification,
Figure BDA0003150762790000051
is a matrix
Figure BDA0003150762790000052
The k-th row and the p-th column.
Compared with the prior art, the invention has the beneficial effects that:
(1) the invention approaches the nonlinear temperature drift characteristic line in the form of a multi-scale filter, and can effectively ensure the accuracy and stability of the approach result;
(2) the invention organically integrates the heterogeneity correction target function into the infrared image splicing treatment, and can simultaneously carry out multi-scale temperature drift characteristic approximation calculation and heterogeneity correction treatment;
(3) the method can effectively remove temperature drift noise in the infrared image, furthest maintain high-frequency information of the image, obtain a high-quality infrared splicing result image, and has higher definition and target identification degree of the target.
Drawings
FIG. 1 is a flowchart of infrared image stitching correction according to the present invention;
FIG. 2 is an initial stitching map of infrared images X and Y to be stitched according to the present invention;
FIG. 3 is a graph of the initial temperature drift of the present invention;
FIG. 4 is a graph of temperature drift after optimization according to the present invention;
FIG. 5 is a graph of infrared image stitching results.
Detailed Description
The invention is further illustrated by the following examples.
The invention provides an infrared image splicing correction method based on temperature drift characteristic line approximation, which can effectively ensure the global optimality and stability of splicing results and can furthest keep the high-frequency information performance of images to be better.
An infrared image stitching correction method, as shown in fig. 1, specifically includes the following steps:
step one, setting an initialization scale parameter d to be 2 and enabling a correction error e of a first iteration to be1Is 1.
And step two, inputting the infrared images X in m rows and n columns and the infrared images Y in s rows and n columns, as shown in FIG. 2. Calculating an extended image of the infrared image X
Figure BDA0003150762790000053
Calculating an extended image of the infrared image Y
Figure BDA0003150762790000054
Is a matrix with m rows and n +2d-2 columns;
Figure BDA0003150762790000055
is a matrix with s rows and n +2d-2 columns; wherein the image is expanded
Figure BDA0003150762790000056
Satisfies the following iterative formula:
Figure BDA0003150762790000057
wherein j is 1,2, … …, m; t-1, 2, … …, d-1;
Figure BDA0003150762790000061
wherein j is 1,2, …, m; t-1, 2, … …, d-1;
Figure BDA0003150762790000062
wherein j is 1,2, … …, m; t is 1,2, … …, n;
expanding an image
Figure BDA0003150762790000063
Satisfies the following iterative formula:
Figure BDA0003150762790000064
wherein j is 1,2, … …, s; t-1, 2, … …, d-1;
Figure BDA0003150762790000065
wherein j is 1,2, … …, s; t-1, 2, … …, d-1;
Figure BDA0003150762790000066
wherein j is 1,2, … …, s; t is 1,2, … …, n.
Step three, calculating a multi-scale filter H and a temperature drift initial curve V; the multi-scale filter H is a 2d-1 dimensional vector, and specifically comprises the following steps:
H＝[12……d*(d-1)……1]/(d+d*(d-1))
the initial curve V of the temperature drift is:
V＝[V1 V2…Vn+2d-2]
in the formula, Xk,pThe value of the k row and the p column in the matrix X;
Yk,pthe values of the k row and the p column in the matrix Y are shown;
Figure BDA0003150762790000067
is a matrix
Figure BDA0003150762790000068
The value of the kth row and the value of the pth column;
Figure BDA0003150762790000069
is a matrix
Figure BDA00031507627900000610
The value of the kth row and the value of the pth column;
Hkis the kth element in vector H;
Vkis the kth element in vector V;
Figure BDA00031507627900000611
the initial curve V of the temperature drift is shown in FIG. 3.
Step four, according to the extended image
Figure BDA00031507627900000612
Expanding an image
Figure BDA00031507627900000613
Carrying out approximation correction optimization processing on the multi-scale filter H and the temperature drift initial curve V to obtain a final correction error ed(ii) a The specific approach of the approximation correction optimization processing is as follows:
s41, setting the iteration number as a, and setting the correction error r of the 0 th iteration0Is 1.
S42, updating and calculating a temperature drift filtering parameter matrix U of the row a and the column n +2 d-2; the calculation method of the temperature drift filtering parameter matrix U comprises the following steps:
when a is 1, initializing, making U in 1 st row and j th column of matrix U1,j＝Vj，j＝1，2，……，n+2d-2；
Updating U1,jThe calculation formula of (2) is as follows:
Figure BDA0003150762790000071
wherein j is d, d +1, … …, d + n-1;
in the formula of Uk,pThe values of the k row and the p column in the matrix U are shown;
Hkis the kth element in vector H;
Vkis the kth element in vector V;
when a is more than 1, initializing the Ua,j＝Ua-1,j，j＝1，2，……，n+2d-2；
Updating Ua,jThe calculation formula of (2) is as follows:
Figure BDA0003150762790000072
s43, calculating an approximation graph Z of S +1 rows and n columns and a corrected error r of the a-th iterationa(ii) a The approximation graph Z satisfies:
Z1,j＝Xm,jwherein j is 1,2, … …, n
Figure BDA0003150762790000073
Wherein t is 1,2, … …, s; j is 1,2, … …, n;
corrected error r for iteration aaComprises the following steps:
Figure BDA0003150762790000074
in the formula, Zk,pIs the value of the kth row and the pth column in the matrix Z.
S44, correcting error r for a (a) th iterationaIs judged when r isa＜ra-1In time, the temperature drift optimization curve under the current scale parameter d
Figure BDA0003150762790000075
Updating the element to the a-th row of the temperature drift filtering parameter matrix U, and adding raIs recorded as the final correctionError edGo to S45; when r isa≥ra-1When the update is not performed, the process proceeds to S45.
S45, p | ra-ra-1| making a judgment when | ra-ra-1|＜1.0×10-5Then, entering the step five; otherwise, let a be a +1, return to S42.
Step five, pair edMaking a judgment when ed＜ed-1The temperature drift curve of the d iteration
Figure BDA0003150762790000076
Record as the temperature drift optimization curve
Figure BDA0003150762790000077
And the d value is recorded as the optimal scale
Figure BDA0003150762790000078
Entering the step six; when e isd≥ed-1And (5) directly entering the step six.
Step six, setting a threshold value D, wherein the value range of the threshold value D is as follows: d is more than or equal to 12 and less than or equal to 16. Comparing the value of D with a threshold value D, and entering a seventh step when D is larger than D; otherwise, returning to the step two;
step seven, optimizing a curve e according to the temperature driftd＜ed-1Optimal size of the optical fiber
Figure BDA0003150762790000081
Carrying out approximate splicing processing on the infrared image X and the infrared image Y to obtain a spliced result graph
Figure BDA0003150762790000082
As shown in fig. 5. Stitched result graph
Figure BDA0003150762790000083
Satisfies the following conditions:
Figure BDA0003150762790000084
wherein i is 1,2, … …, m; j-1, 2, … …，n；
Figure BDA0003150762790000085
Wherein i is 1,2, … …, s; j is 1,2, … …, n;
in the formula (I), the compound is shown in the specification,
Figure BDA0003150762790000086
is a matrix
Figure BDA0003150762790000087
The k-th row and the p-th column.
The following is further illustrated with reference to specific examples:
taking a 12-bit infrared image X of 528 × 4101 and a 12-bit infrared image Y of 545 × 4101 to be stitched as an example, a flow of an infrared image stitching correction method based on temperature drift characteristic line approximation is shown in fig. 1, and the specific process is as follows:
(1) setting the initialization scale parameter d to 2, correcting the error e1＝1；
(2) Calculating an extended image according to the input m rows and n columns of infrared images X and the input s rows and n columns of infrared images Y
Figure BDA0003150762790000088
And
Figure BDA0003150762790000089
calculating a multi-scale filter H and a temperature drift initial curve V; in this embodiment, as shown in fig. 2, the input infrared image X to be stitched is a 528 × 4101 matrix, and the stitched infrared image Y is a 545 × 4101 matrix;
Figure BDA00031507627900000810
wherein j is 1,2, … …, m; t-1, 2, … …, d-1;
Figure BDA00031507627900000811
wherein j is 1,2, …, m; t-1, 2, … …, d-1;
Figure BDA00031507627900000812
wherein j is 1,2, … …, m; t is 1,2, … …, n;
expanding an image
Figure BDA00031507627900000813
Satisfies the following iterative formula:
Figure BDA00031507627900000814
wherein j is 1,2, … …, s; t-1, 2, … …, d-1;
Figure BDA00031507627900000815
wherein j is 1,2, … …, s; t-1, 2, … …, d-1;
Figure BDA00031507627900000816
wherein j is 1,2, … …, s; t is 1,2, … …, n;
h is a 2d-1 dimensional vector, H ═ 12 … … d (d-1) … … 1]/(d + d (d-1));
V＝[V1 V2…Vn+2d-2]，
Figure BDA00031507627900000817
the initial temperature drift curve V is shown in fig. 3.
(3) From expanded images
Figure BDA0003150762790000091
And
Figure BDA0003150762790000092
carrying out approximation correction optimization processing on the multi-scale filter H and the temperature drift initial curve V to obtain a final correction error ed(ii) a The method comprises the following specific steps:
(3-1) settingThe iteration number a is 1, and the correction error r of the 0 th iteration is set0Is 1;
(3-2) updating and calculating a temperature drift filtering parameter matrix U of the a-th row and the n +2d-2 th column;
when a is 1, initializing, making U in 1 st row and j th column of matrix U1,j＝Vj，j＝1，2，……，n+2d-2；
Then, the calculation formula is updated as follows:
Figure BDA0003150762790000093
wherein: u shapek,pThe values of the k row and the p column in the matrix U are shown;
Hkis the kth element in vector H;
Vkis the kth element in vector V;
if a is>1 hour, initializing Ua,j＝Ua-1,j，j＝1，2，…,n+2d-2；
Then, the calculation formula is updated as follows:
Figure BDA0003150762790000094
(3-3) calculating the s +1 row and n column approximation graph Z and the corrected error r of the a-th iterationa；
Z1,j＝Xm,jWherein j is 1,2, …, n;
Figure BDA0003150762790000095
wherein t is 1,2, …, s; j is 1,2, …, n;
Figure BDA0003150762790000096
wherein: zk,pIs the value of the kth row and the pth column in the matrix Z.
(3-4) if ra＜ra-1Optimizing the temperature drift curve under the current scale parameter d
Figure BDA0003150762790000097
Updating the element to the a-th row of the temperature drift filtering parameter matrix U, and adding raIs expressed as the final correction error edEntering (3-5); when r isa≥ra-1When the data is not updated, the data is directly entered into (3-5).
(3-5) when | ra-ra-1|＜1.0×10-5If so, the step (4) is entered, otherwise, the step (3-2) is returned to the step a + 1.
(4) If ed＜ed-1Temperature drift curve of the d-th iteration
Figure BDA0003150762790000098
Record as the temperature drift optimization curve
Figure BDA0003150762790000099
And the d value is recorded as the optimal scale
Figure BDA00031507627900000910
And entering the step (5); when e isd≥ed-1And (5) directly entering the step (5).
(5) And (3) setting a threshold value D, if D is more than or equal to 12 and less than or equal to 16, entering the step (6), and if D is equal to D +1, returning to the step (2). The resulting optimized temperature drift curve is shown in FIG. 4.
(6) Temperature drift optimization curve obtained by optimization processing
Figure BDA0003150762790000101
And an optimal scale
Figure BDA0003150762790000102
Carrying out approximate splicing processing on the input infrared images X and Y to obtain a spliced result graph
Figure BDA0003150762790000103
The results of the infrared image example non-uniformity correction are shown in fig. 5.
The infrared image splicing correction method based on temperature drift characteristic line approximation provided by the invention has the advantages that the nonlinear temperature drift characteristic line is approximated in the form of the multi-scale filter, the accuracy and the stability of an approximation result can be effectively ensured, meanwhile, the non-uniformity correction target function is organically integrated into the infrared image splicing treatment, the multi-scale temperature drift characteristic approximation calculation and the non-uniformity correction treatment can be simultaneously carried out, the temperature drift noise in the infrared image can be effectively removed, the high-frequency information of the image is kept to the maximum extent, the high-quality infrared splicing result image is obtained, and the definition and the target identification degree of a target are higher.
Although the present invention has been described with reference to the preferred embodiments, it is not intended to limit the present invention, and those skilled in the art can make variations and modifications of the present invention without departing from the spirit and scope of the present invention by using the methods and technical contents disclosed above.

Claims (9)

1. An infrared image stitching correction method based on temperature drift characteristic line approximation is characterized by comprising the following steps: the method comprises the following steps:
step one, setting an initialization scale parameter d and a correction error e of a first iteration1；
Inputting the infrared image X of m rows and n columns and the infrared image Y of s rows and n columns, and calculating the extended image of the infrared image X
Figure FDA0003150762780000011
Calculating an extended image of the infrared image Y
Figure FDA0003150762780000012
Step three, calculating a multi-scale filter H and a temperature drift initial curve V;
step four, according to the extended image
Figure FDA0003150762780000013
Expanding an image
Figure FDA0003150762780000014
Carrying out approximation correction optimization processing on the multi-scale filter H and the temperature drift initial curve V to obtain a final correction error ed；
Step five, pair edMaking a judgment when ed＜ed-1The temperature drift curve of the d iteration
Figure FDA0003150762780000015
Record as the temperature drift optimization curve
Figure FDA0003150762780000016
And the d value is recorded as the optimal scale
Figure FDA0003150762780000017
Entering the step six; when e isd≥ed-1Directly entering the step six;
step six, setting a threshold value D, comparing the value D with the threshold value D, and entering step seven when D is larger than D; otherwise, returning to the step two;
step seven, optimizing a curve e according to the temperature driftd＜ed-1Optimal size of the optical fiber
Figure FDA0003150762780000018
Carrying out approximate splicing processing on the infrared image X and the infrared image Y to obtain a spliced result graph
Figure FDA0003150762780000019
2. The infrared image stitching correction method based on temperature drift characteristic line approximation as claimed in claim 1, characterized in that: in the first step, d is 2, e1＝1。
3. The infrared image stitching correction method based on temperature drift characteristic line approximation as claimed in claim 2, characterized in that: in the second step, the first step is carried out,
Figure FDA00031507627800000110
is a matrix with m rows and n +2d-2 columns;
Figure FDA00031507627800000111
is a matrix with s rows and n +2d-2 columns; wherein the image is expanded
Figure FDA00031507627800000112
Satisfies the following iterative formula:
Figure FDA00031507627800000113
wherein j is 1,2, … …, m; t-1, 2, … …, d-1;
Figure FDA00031507627800000114
wherein j is 1,2, …, m; t-1, 2, … …, d-1;
Figure FDA00031507627800000115
wherein j is 1,2, … …, m; t is 1,2, … …, n;
expanding an image
Figure FDA00031507627800000116
Satisfies the following iterative formula:
Figure FDA0003150762780000021
wherein j is 1,2, … …, s; t-1, 2, … …, d-1;
Figure FDA0003150762780000022
wherein j is 1,2, … …, s; t-1, 2, … …, d-1;
Figure FDA0003150762780000023
wherein j is 1,2, … …, s; t is 1,2, … …, n.
4. The infrared image stitching correction method based on temperature drift characteristic line approximation as claimed in claim 3, characterized in that: in the third step, the multi-scale filter H is a 2d-1 dimensional vector, and specifically comprises the following steps:
H＝[12……d*(d-1)……1]/(d+d*(d-1))
the initial curve V of the temperature drift is:
V＝[V1 V2 … Vn+2d-2]
in the formula, Xk,pThe value of the k row and the p column in the matrix X;
Yk,pthe values of the k row and the p column in the matrix Y are shown;
Figure FDA0003150762780000024
is a matrix
Figure FDA0003150762780000025
The value of the kth row and the value of the pth column;
Figure FDA0003150762780000026
is a matrix
Figure FDA0003150762780000027
The value of the kth row and the value of the pth column;
Hkis the kth element in vector H;
Vkis the kth element in vector V;
Figure FDA0003150762780000028
5. the infrared image stitching correction method based on temperature drift characteristic line approximation as claimed in claim 4, characterized in that: in the fourth step, a specific approach to the optimization processing of correction is as follows:
s41, setting the iteration number as a, and setting the correction error r of the 0 th iteration0Is 1;
s42, updating and calculating a temperature drift filtering parameter matrix U of the row a and the column n +2 d-2;
s43, calculating an approximation graph Z of S +1 rows and n columns and a corrected error r of the a-th iterationa；
S44, correcting error r for a (a) th iterationaIs judged when r isa＜ra-1In time, the temperature drift optimization curve under the current scale parameter d
Figure FDA0003150762780000029
Updating the element to the a-th row of the temperature drift filtering parameter matrix U, and adding raIs expressed as the final correction error edGo to S45; when r isa≥ra-1When the update is not performed, the flow proceeds directly to S45;
s45, p | ra-ra-1I, judging when ra-ra-1|＜1.0×10-5Then, entering the step five; otherwise, let a be a +1, return to S42.
6. The infrared image stitching correction method based on temperature drift characteristic line approximation as claimed in claim 5, characterized in that: in S42, the calculation method of the temperature drift filtering parameter matrix U is:
when a is 1, initializing, making U in 1 st row and j th column of matrix U1,j＝Vj，j＝1，2，……，n+2d-2；
Updating U1,jThe calculation formula of (2) is as follows:
Figure FDA0003150762780000031
wherein j is d, d +1, … …, d + n-1;
in the formula of Uk,pThe values of the k row and the p column in the matrix U are shown;
Hkis the kth element in vector H;
Vkis the kth element in vector V;
when a is more than 1, initializing the Ua,j＝Ua-1,j，j＝1，2，……，n+2d-2；
Updating Ua,jThe calculation formula of (2) is as follows:
Figure FDA0003150762780000032
7. the infrared image stitching correction method based on temperature drift characteristic line approximation as claimed in claim 5, characterized in that: in S43, the approximation map Z satisfies:
Z1,j＝Xm,jwherein j is 1,2, … …, n
Figure FDA0003150762780000033
Wherein t is 1,2, … …, s; j is 1,2, … …, n;
corrected error r for iteration aaComprises the following steps:
Figure FDA0003150762780000034
in the formula, Zk,pIs the value of the kth row and the pth column in the matrix Z.
8. The infrared image stitching correction method based on temperature drift characteristic line approximation as claimed in claim 1, characterized in that: in the sixth step, the value range of the threshold D is as follows: d is more than or equal to 12 and less than or equal to 16.
9. The infrared image stitching correction method based on temperature drift characteristic line approximation as claimed in claim 1, characterized in that: in the seventh step, the spliced result graph
Figure FDA0003150762780000035
Satisfies the following conditions:
Figure FDA0003150762780000036
wherein i is 1,2, … …, m; j is 1,2, … …, n;
Figure FDA0003150762780000037
wherein i is 1,2, … …, s; j is 1,2, … …, n;
in the formula (I), the compound is shown in the specification,
Figure FDA0003150762780000038
is a matrix
Figure FDA0003150762780000039
The k-th row and the p-th column.
CN202110763276.5A 2021-07-06 2021-07-06 Infrared image stitching correction method based on temperature drift characteristic line approximation Active CN113658054B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110763276.5A CN113658054B (en) 2021-07-06 2021-07-06 Infrared image stitching correction method based on temperature drift characteristic line approximation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110763276.5A CN113658054B (en) 2021-07-06 2021-07-06 Infrared image stitching correction method based on temperature drift characteristic line approximation

Publications (2)

Publication Number Publication Date
CN113658054A true CN113658054A (en) 2021-11-16
CN113658054B CN113658054B (en) 2024-03-29

Family

ID=78489132

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110763276.5A Active CN113658054B (en) 2021-07-06 2021-07-06 Infrared image stitching correction method based on temperature drift characteristic line approximation

Country Status (1)

Country Link
CN (1) CN113658054B (en)

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040042685A1 (en) * 2002-08-28 2004-03-04 Lingxiang Zhou Image warping correction in forming 360 degree panoramic images
CN102042878A (en) * 2010-10-21 2011-05-04 电子科技大学 Infared nonuniformity correction method for removing temperature shift
JP2012058064A (en) * 2010-09-08 2012-03-22 Nippon Ceramic Co Ltd Method for correcting data of infrared imaging device
CN106197690A (en) * 2016-06-27 2016-12-07 湖北久之洋红外系统股份有限公司 Image calibrating method under the conditions of a kind of wide temperature range and system
CN106875374A (en) * 2016-12-21 2017-06-20 北京空间机电研究所 A kind of Weak link image splicing method based on line feature
CN108240863A (en) * 2016-12-23 2018-07-03 南京理工大学 For real-time 2 asymmetric correction methods of Uncooled infrared camera
CN110689485A (en) * 2019-10-14 2020-01-14 中国空气动力研究与发展中心超高速空气动力研究所 SIFT image splicing method applied to infrared nondestructive testing of large pressure container
CN111080567A (en) * 2019-12-12 2020-04-28 长沙理工大学 Remote sensing image fusion method and system based on multi-scale dynamic convolution neural network
CN112085803A (en) * 2020-07-27 2020-12-15 北京空间机电研究所 Multi-lens multi-detector splicing type camera color consistency processing method

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040042685A1 (en) * 2002-08-28 2004-03-04 Lingxiang Zhou Image warping correction in forming 360 degree panoramic images
JP2012058064A (en) * 2010-09-08 2012-03-22 Nippon Ceramic Co Ltd Method for correcting data of infrared imaging device
CN102042878A (en) * 2010-10-21 2011-05-04 电子科技大学 Infared nonuniformity correction method for removing temperature shift
CN106197690A (en) * 2016-06-27 2016-12-07 湖北久之洋红外系统股份有限公司 Image calibrating method under the conditions of a kind of wide temperature range and system
CN106875374A (en) * 2016-12-21 2017-06-20 北京空间机电研究所 A kind of Weak link image splicing method based on line feature
CN108240863A (en) * 2016-12-23 2018-07-03 南京理工大学 For real-time 2 asymmetric correction methods of Uncooled infrared camera
CN110689485A (en) * 2019-10-14 2020-01-14 中国空气动力研究与发展中心超高速空气动力研究所 SIFT image splicing method applied to infrared nondestructive testing of large pressure container
CN111080567A (en) * 2019-12-12 2020-04-28 长沙理工大学 Remote sensing image fusion method and system based on multi-scale dynamic convolution neural network
CN112085803A (en) * 2020-07-27 2020-12-15 北京空间机电研究所 Multi-lens multi-detector splicing type camera color consistency processing method

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
ZHENHUA LI 等: "Spatially adaptive hybrid variational model for temperature-dependent nonuniformity correction of infrared images", 2020 SPIE, vol. 59, no. 12 *
颜佩琴 等: "基于温漂自动补偿的高稳定性Ｓｉ－ＡＰＤ单光子探测器", 激光与光电子学进展, vol. 54, no. 8 *

Also Published As

Publication number Publication date
CN113658054B (en) 2024-03-29

Similar Documents

Publication Publication Date Title
CN103528694B (en) A kind of method of temperature of infrared heat image instrument measuring target object
CN109272520B (en) Self-adaptive infrared focal plane non-uniform correction method combining motion guidance and edge detection
CN110766058B (en) Battlefield target detection method based on optimized RPN (resilient packet network)
CN115752757A (en) On-orbit non-uniformity correction method for infrared scanning camera with separated detector and inner black body
CN113705788A (en) Infrared image temperature estimation method and system based on full convolution neural network
CN110555231A (en) Dynamic simulation model correction method
CN112393807A (en) Infrared image processing method, device, system and computer readable storage medium
CN110006529B (en) Output correction method and device for infrared detection device
US10861187B2 (en) Method of processing object detection data
CN109242832B (en) Multi-temporal multispectral remote sensing image change detection method and system
Huang et al. Object-based attention mechanism for color calibration of UAV remote sensing images in precision agriculture
CN113658054A (en) Infrared image splicing correction method based on temperature drift characteristic line approximation
CN114187189A (en) Aircraft multispectral image radiation consistency correction method
CN107993193B (en) Tunnel lining image splicing method based on illumination equalization and surf algorithm improvement
CN112150377B (en) Infrared image non-uniformity correction coefficient analysis solution alternation iteration optimization method
CN111854960A (en) Calibration method for measuring human face temperature based on thermal imager
CN111899284A (en) Plane target tracking method based on parameterized ESM network
CN116342660B (en) Multi-scale analysis fusion weighting filtering bionic compound eye optical flow field estimation method
CN117115488B (en) Water meter detection method based on image processing
CN117091708A (en) Infrared detector response characteristic drift evaluation method based on time sequence analysis
CN111079838B (en) Hyperspectral band selection method based on double-flow-line low-rank self-expression
CN116681721B (en) Linear track detection and tracking method based on vision
CN117197651B (en) Method and system for extracting field by combining edge detection and semantic segmentation
CN103229497B (en) For the method and apparatus for the screen window effect for estimating image detection device
CN112989714B (en) Training method and device for detector adjustment model

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant