CN105527072B - A method of the in-orbit focal length of remote sensor is obtained based on remote sensing images - Google Patents
A method of the in-orbit focal length of remote sensor is obtained based on remote sensing images Download PDFInfo
- Publication number
- CN105527072B CN105527072B CN201410520730.4A CN201410520730A CN105527072B CN 105527072 B CN105527072 B CN 105527072B CN 201410520730 A CN201410520730 A CN 201410520730A CN 105527072 B CN105527072 B CN 105527072B
- Authority
- CN
- China
- Prior art keywords
- focal length
- opposite edges
- response
- image
- remote sensor
- 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.)
- Active
Links
Landscapes
- Image Analysis (AREA)
Abstract
The present invention provides a kind of method for obtaining the in-orbit focal length of remote sensor based on remote sensing images, comprising the following steps: by the in-orbit imaging of computer simulation remote sensor, emulation obtains focal length emulating image of the multiple groups under different focal length grade；Calculate the opposite edges response of each width focal length emulating image；It is for statistical analysis to opposite edges response, the data of interference modeling are removed, then using opposite edges response as variable, using remote sensor focal length as dependent variable, regression modeling is carried out, obtains the mathematical model between image opposite edges response and remote sensor focal length；For the remote sensing images to be processed of acquisition, its opposite edges response parameter is calculated first, and to calculate resulting opposite edges response as input, the mathematical model of foundation is inputted, to obtain the focal length of remote sensor.
Description
Technical field
The present invention relates to remote sensing satellite image data processing fields, distant based on remote sensing images acquisition in particular to one kind
The method of the in-orbit focal length of sensor.
Background technique
For ordinary lens, focal length is just focal point to the distance of lens centre, and for optical sensor, it is burnt
Away from being with the equivalent focal length of the optical system (EFL, effective focal length) come what is indicated.The optical system is often
It is made of several lens and reflecting mirror, each lens and reflecting mirror have respective focus and focal length, to entire optical system
For, it has the combined focal and combined focal length of a pair of of system, referred to as " equivalent focus " and " equivalent focal length ".In-orbit remote sensor is burnt
Away from variation will result directly in the difference of imaging results, cause the variation of picture quality.
Remote sensor is since long-time is in space, so the problem of saving bit by bit the machinery to get off and circuit aspect will lead to
Optical system focal length changes, this just will directly cause the variation of the quality of remote sensing images captured by it, such as defocus meeting
It generates fuzzy.This just needs the staff on ground according to the imaging contexts of remote sensor regularly inspection result, manual adjustment it
Focal length, to make up because of reasons such as long-time mechanical agings caused by focal length variations problem.This adjustment mode needs to have experience
Remote sensing expert according to the observation multiple image as a result, by virtue of experience make adjustment appropriate to the focal length of remote sensor, and
When adjustment, it is also necessary to generate the quality condition of image again according to remote sensor to judge whether this adjustment succeeds.This side
Method had not only consumed largely manually, but also needed certain imaging cycle to judge the reasonability of adjustment result.
So being badly in need of a kind of method that can instruct the automatic in-orbit adjustment focal length of remote sensor.But remote sensor focal-distance tuning range
Acquisition be faced with very big challenge, staff lacks the correct amount of effective approach focal length measurement variation on the ground.
Summary of the invention
It is an object of the invention to can not solve automatic accurate acquisition remote sensor in view of the prior art because mechanical aging is sent out
Raw focal length variations problem proposes a kind of method for obtaining the in-orbit focal length of remote sensor based on remote sensing images.
Above-mentioned purpose of the invention realizes that dependent claims are to select else or have by the technical characteristic of independent claims
The mode of benefit develops the technical characteristic of independent claims.
To reach above-mentioned purpose, technical solution of the present invention is as follows:
A method of the in-orbit focal length of remote sensor is obtained based on remote sensing images, comprising the following steps:
Step 1 passes through the in-orbit imaging of computer simulation remote sensor, and emulation obtains focal length of the multiple groups under different focal length grade
Emulating image；
Step 2, the opposite edges response for calculating each width focal length emulating image；
Step 3, the opposite edges response that step 2 is calculated are for statistical analysis, remove the data of interference modeling,
Then using the image opposite edges response after statisticalling analyze as variable, using remote sensor focal length as dependent variable, regression modeling is carried out,
Obtain the mathematical model between image opposite edges response and remote sensor focal length；And
Step 4: for the remote sensing images to be processed of acquisition, calculating its opposite edges response parameter, and first to calculate
The opposite edges response obtained inputs the mathematical model that the step 3 is established, to obtain the focal length of remote sensor as input.
From the above technical solution of the present invention shows that, proposed by the invention to obtain the in-orbit coke of remote sensor based on remote sensing images
Away from method, compared with prior art, remarkable result is:
Using method proposed by the present invention, the focal length parameter of remote sensor can be estimated according to the mass parameter of image, for referring to
It leads remote sensor and is automatically completed in-orbit Focussing.Remote sensing expert is required to the adjustment of the in-orbit parameter of remote sensor in the past relatively
It rule of thumb completes, using method of the invention come founding mathematical models, and opposite edges response is calculated according to image
Value, so that it may the focal length of remote sensor is finally inversed by, so that remote sensor be instructed to make corresponding adjustment to obtain the remote sensing figure of high-quality
Picture, for remote sensor it is in-orbit carry out focal length calibration provide theoretical foundation and technical support, while greatly reduce manpower and when
Between cost.
Detailed description of the invention
Fig. 1 is that the implementation process of method of an embodiment of the present invention based on the remote sensing images acquisition in-orbit focal length of remote sensor is shown
It is intended to.
Fig. 2 a-2c is three examples of target image.
Fig. 3 a-3c is the example of three focal length emulating images of corresponding diagram 2a-2c.
Fig. 4 is an example of Sobel operator.
Fig. 5 is the example of edge image.
Fig. 6 a is that the edge image of Fig. 5 is added to the example after selecting frame；Fig. 6 b is plus selecting the former focal length analogous diagram after frame
The example of picture.
Fig. 7 a is the scatter plot of the skirt response in the direction y；Fig. 7 b is the fitted figure using S type curve.
Fig. 8 a is the scatter plot of the focal length and skirt response before normalization；Fig. 8 b is that the focal length after normalization is corresponding to edge
Scatter plot.
Fig. 9 is the matched curve schematic diagram of focal length and opposite edges response.
Specific embodiment
In order to better understand the technical content of the present invention, special to lift specific embodiment and institute's accompanying drawings is cooperated to be described as follows.
On the whole, the method for obtaining the in-orbit focal length of remote sensor based on remote sensing images that the present embodiment proposes, passes through calculating
Machine simulates the image generated of remote sensor under different focal length, then from the opposite edges response of image, for multiple etc.
The opposite edges response of the emulating image of grade focal length carries out regression analysis, establishes the mathematical model of focal length and opposite edges response,
To which achieved the purpose that can be by image inverting focal length.It, i.e., can be distant using its guidance after having obtained the focal length of remote sensor
Sensor is automatically adjusted to suitable focal length.
As shown in Figure 1, preferred embodiment according to the present invention, a kind of to obtain the in-orbit focal length of remote sensor based on remote sensing images
Method, comprising the following steps:
Step 1 passes through the in-orbit imaging of computer simulation remote sensor, and emulation obtains focal length of the multiple groups under different focal length grade
Emulating image；
Step 2, the opposite edges response for calculating each width focal length emulating image；
Step 3, the opposite edges response that step 2 is calculated are for statistical analysis, remove the data of interference modeling,
Then using the image opposite edges response after statisticalling analyze as variable, using remote sensor focal length as dependent variable, regression modeling is carried out,
Obtain the mathematical model between image opposite edges response and remote sensor focal length；And
Step 4: for the remote sensing images to be processed of acquisition, calculating its opposite edges response parameter, and first to calculate
The opposite edges response obtained inputs the mathematical model that the step 3 is established, to obtain the focal length of remote sensor as input.
Below with reference to shown in Fig. 2-Fig. 9, the specific implementation of above steps is described in detail.
Step 1 obtains focal length emulating image
As previously mentioned, by the in-orbit imaging of computer simulation remote sensor, emulation obtains multiple groups in different cokes in the present embodiment
Away from the focal length emulating image under grade.
Using P width image of a certain satellite in different latitude as target image, then according to modulation transfer function model
And ray trace model, by the imaging process of the visual remote sensing device of the computer simulation satellite in orbit, emulation is obtained
P*Q width focal length emulating image under different focal length grade Q.
For example, with P of No. three satellites of resource in different latitude, (P value is the bigger the better, and P value is bigger, finally established model
It is more representative) width image is as target image, and P is taken as 17 in the present embodiment.Then distant by computer simulation
The in-orbit imaging of sensor obtains the focal length emulating image of different focal length grade Q.Specifically, according to existing modulation transfer function mould
Type and ray trace model, to simulate the imaging process of the visual remote sensing device of No. three satellites of resource in orbit, emulation is obtained
17 width images.
As optional embodiment, in this example, being provided with focal length grade Q value is that 39 (from 60 to 1960, spacing is
50, unit is millimeter), naturally it is also possible to it is other values.The value of focal length grade Q is corresponding to be used when being regression modeling
Number of samples, in general value are relatively mild 30 to 50.
39 width focal length emulating images can be obtained by every width target image to give respectively as shown in Fig. 2 a-2c and Fig. 3 a-3c
The example of part target image and partial simulation image is gone out.
Step 2, the opposite edges response for calculating image
Whether clearly opposite edges response is a measurement image border important parameter, and the readability of image border
Largely reflect the quality of piece image.The key for calculating opposite edges response is the construction of edge response curve,
So-called edge response curve is the curve that edge line two sides pixel value changes in image, and opposite edges response is then in the side x
To on the direction y, the geometric average of the difference of pixel value on each 0.5 pixel in edge line two sides.
In the present embodiment, carry out the opposite edges response of computer sim- ulation image in the following way.
1) edge detection is carried out to focal length emulating image above-mentioned using Sobel operator, then carries out threshold process, obtain side
Edge image.
Fig. 4 show the example of a Sobel operator.Edge detection is carried out using the Sobel operator, is being carried out at threshold value
It manages, in the present embodiment, threshold value is ordinarily selected to 33% of maximum pixel value in image.
By Sobel operator, treated that image is not the ideal edge image of a width because in figure pixel value change
Changing range is still 0-255, does not pass through binary conversion treatment, and after threshold process, greater than the pixel value of given threshold
255 are set to, the pixel value less than threshold value is set to 0, so the image after threshold process is that a width only has " black " and " white "
Edge image.
Obtained edge image corresponds to shown in Fig. 2 a and Fig. 3 a, an example of edge image, as shown in Figure 5.
2) target area of the continuous block in an edge as construction edge response curve then, is selected in edge image,
And outlined using box, box as shown in Figure 6 a outlines region, and Fig. 6 b is shown plus the original image for selecting frame.
3) to selecting in frame the i.e. marginal point of target area using least square method to carry out straight line fitting, will fitting obtain it is straight
Object edge line of the line as the width image, such as the straight line in the target area Fig. 6 b；Then calculating selects all pixels point in frame to arrive
Distance on the direction x and the direction y of object edge line, the equation of object edge line are as follows:
Y=ax+b (1)
Wherein
For any pixel point (x0,y0), to the distance in the direction x of object edge line are as follows:
To the distance in the direction y of object edge line are as follows:
distancey=y0-(ax0+b) (5)
Then, using pixel to the distance in the object edge direction line x or the direction y as abscissa, pixel value is as vertical
Coordinate constructs the scatter plot of skirt response, such as Fig. 7 a.
4) Function Modules of edge response curve are constructed for the considerations of carrying out batch processing to image, in the present embodiment
Type.Edge response curve is equivalent to a S type curve, is fitted with the S type curve.
In the present embodiment, use function prototype forCurve be fitted, to its specific form add four control
Parameter a processed1,a2,a3,a4, to form following S type curvilinear function models:
The curve matching function of reusing Matlab is fitted above-mentioned function, is obtaining parameter a1,a2,a3,a4It
Afterwards, the construction of edge response curve is finally completed.
Fig. 7 b show the fitted figure using S type curve, i.e. image in Fig. 2 a uses the fitting of the S type curve of formula 6
Figure.
5) under the premise of obtaining edge response curve, opposite edges analog value RER is calculated based on following formulas (7):
Wherein:
ERxFor the opposite edges analog value of X-direction, ERyFor the opposite edges analog value of Y-direction.
As it can be seen that in this method, selecting the target of each width target image construction edge response curve first in this step 2
Then region is responded using the opposite edges that the above method calculates each width emulating image, to constitute needed for modeling in next step
The data wanted.
Step 3, regression modeling
Since the terrestrial object information of 17 width target images is there is very big difference, the contrast of many images be not it is very strong,
The grey value difference of object edge line two sides can be seriously affected, this Value levels for resulting in the opposite edges of different images to respond
There is very big fluctuation.In order to eliminate these influences, need to do normalized pre- place to the resulting opposite edges response of step 2
Reason.
In the present embodiment, it is normalized using following formulas (9):
Wherein, xijIndicate the opposite edges response of the jth width emulating image of the i-th width target image, i=1,2 ..., P
(it has been observed that P is taken as 17 in this example), j=1,2 ..., n, n correspond to the aforementioned focal length grade Q (at this value as 39) set.Figure
8a and 8b is the scatter plot of normalization front and back focal length and opposite edges response respectively, it can be seen that every group of emulating image opposite edges
The variation tendency of response is almost the same, and after normalization, data are more regular, is easy to the modeling processing of lower section.
It in the present embodiment, is analyzed using SPSS software (i.e. statistical product and service solution software), analysis becomes
Correlation between amount (i.e. between opposite edges response and focal length), to remove the data of interference modeling.
Specifically, judged using the correlation arbitration functions of SPSS software itself, if variable Normal Distribution,
Correlation judgement is carried out using Pearson came (Pearson) related coefficient, if variable disobeys normal distribution, uses Spearman
(Spearman) related coefficient carries out correlation judgement.
And SPSS software itself provides the non-parametric test method of two kinds of normal distributions: Kolmogorov-Smirnov inspection
It tests and is examined with Shapiro-Wilk.(5000 are less than) in the case where sample number is less, and Shapiro-Wilk inspection has more can
Reliability.It can consider that variable is Normal Distribution when the level of signifiance is greater than 0.01.
Judge by SPSS software, normal distribution is all disobeyed in the opposite edges response of all focal length emulating images, therefore
In the present embodiment, using Spearman related coefficient come the correlation between test variable, following table lists 17 groups of analogous diagrams of focal length
The Spearman related coefficient responded as opposite edges:
The Spearman related coefficient of 1 focal length of table and opposite edges response
Through the correlation analysis to opposite edges response and focal length it is found that in 17 groups of images of focal length emulation, only
The absolute value of the Spearman related coefficient of 6th group of emulating image opposite edges response is less than 0.95, and level of signifiance sig=
0.02 > 0.01, this just illustrates that focal length and the correlation of the opposite edges response of this group of image be not strong, so removing this group of image.
Next, modeled to the data after rejecting, using opposite edges response focal length each degradation level
Average value i.e. in focal length grade is as the final data for participating in curve matching, using the opposite edges response after average as change
Amount, is carried out curve fitting using focal length as dependent variable, obtains a variety of possible functions using the curve estimation of SPSS software itself
Model and its fitting degree assess Value Data, and wherein the maximum function model of fitting degree assessed value is the most finally established for selection
The collective model of opposite edges response and focal length.
Remove the image removed, there remains 16 groups of focal length emulating images and participate in modeling, every group of image can be formed one group it is complete
The corresponding relationship of whole opposite edges response and focal length, and theoretically every group of emulating image can construct a mathematical model, be
A unified model is obtained, therefore (i.e. burnt in focal length each degradation level using opposite edges response in the present embodiment
Away from grade) on average value as the final data for participating in curve matching.Because in each degradation level of focal length, data
Fluctuation is all very little, and the standard deviation of data in partial deterioration grade, the standard deviation in most of degradation level are listed in table 2
All very littles, so the way being averaged is relatively reasonable.
The standard deviation of emulating image RER in 2 focal length partial deterioration grade of table
The purpose that the present embodiment is proposed is to release focal length by the way that the opposite edges responses of remote sensing images is counter, so will
As variable, focal length carries out curve fitting opposite edges response after average as dependent variable.
In order to evaluate the function model that fitting obtains, coefficient of determination is introduced as criterion.Such as coefficient of determination R2Or it is aobvious
Work property examines F value, and in coefficient of determination R2Or when significance test F value maximum, corresponding function model is selected as most
The function model established afterwards, therefore set up the model between focal length and opposite edges response.
It is the calculation formula of two statistics below:
Wherein p is the number of explanatory variable in regression coefficient equation.
A variety of possible function models can be obtained using the curve estimation function of SPSS software itself, as previously mentioned, and leading to
Cross the coefficient of determination R that SPSS software provides2Or the statistical data of significance test F value, it selectes where it is determined that coefficients R2Or it is aobvious
Work property examines the maximum corresponding function model of F value, as the model between focal length and opposite edges response.
With the aforementioned resulting fitting result of 16 width image, using F value as example, following Table 3 lists the biggish several songs of F value
Line model.
3 curve estimation statistical data of table
As it can be seen that wherein the F value of piecewise function is maximum, therefore choose the piecewise function (segmentation of two cubic functions composition
Function), matched curve is drawn, as shown in Figure 9.
Data are divided into two sections to carry out curve fitting respectively (using RER=0.7506 as boundary), obtain two cubic curve moulds
Type, after the two is stitched together, effect is very ideal.The following are the function models that curve matching obtains:
Step 4, the in-orbit focal length for obtaining remote sensor
According to the model that abovementioned steps 3 obtain, when obtaining one group of remote sensing images, the opposite edges by calculating image are rung
It should be worth, and as input, input the aforementioned model finally selected, exported according to model, obtain the focal length of remote sensor.
Although the present invention has been disclosed as a preferred embodiment, however, it is not to limit the invention.Skill of the present invention
Has usually intellectual in art field, without departing from the spirit and scope of the present invention, when can be used for a variety of modifications and variations.Cause
This, the scope of protection of the present invention is defined by those of the claims.
Claims (3)
1. a kind of method for obtaining the in-orbit focal length of remote sensor based on remote sensing images, which comprises the following steps:
Step 1 passes through the in-orbit imaging of computer simulation remote sensor, and emulation obtains focal length emulation of the multiple groups under different focal length grade
Image；
Step 2, the opposite edges response for calculating each width focal length emulating image；
Step 3, the opposite edges response that step 2 is calculated are for statistical analysis, remove the data of interference modeling, then
Using the image opposite edges response after statisticalling analyze as variable, using remote sensor focal length as dependent variable, regression modeling is carried out, is obtained
Mathematical model between image opposite edges response and remote sensor focal length；And
Step 4: for the remote sensing images to be processed of acquisition, calculating its opposite edges response parameter first, and resulting to calculate
Opposite edges response inputs the mathematical model that the step 3 is established, to obtain the focal length of remote sensor as input；
The realization of the step 1 the following steps are included:
Using P width image of a certain satellite in different latitude as target image, then according to modulation transfer function model and light
Line trace model, by the imaging process of the visual remote sensing device of the computer simulation satellite in orbit, emulation obtains difference
P*Q width focal length emulating image under focal length grade Q；
In the step 2, the calculating of opposite edges response the following steps are included:
1) edge detection is carried out using Sobel operator focusing emulating image, then carries out threshold process, obtain edge image,
Middle threshold value is selected as 33% of maximum pixel value in focal length emulating image；
2) target area of the continuous block in an edge as construction edge response curve is selected in edge image, and uses box
It outlines；
3) straight line fitting is carried out using least square method to all marginal points of selected target area, the straight line that fitting is obtained
Object edge line as the width image；Then all pixels point is calculated in target area to the direction x of the object edge line and y
Distance on direction, the equation of object edge line are as follows:
Y=ax+b (1)
Wherein
For any pixel point (x0,y0), to the distance in the direction x of object edge line are as follows:
To the distance in the direction y of object edge line are as follows:
distancey=y0-(ax0+b) (5)
Then, using pixel to the distance in the object edge direction line x or the direction y as abscissa, pixel value as ordinate,
Construct the scatter plot of skirt response；
4) construct edge response curve function model, edge response curve is equivalent to a S type curve, with the S type curve into
Row fitting, wherein use function prototype forCurve be fitted, to its specific form add four control parameters
a1,a2,a3,a4, to form following S type curvilinear function models:
The curve matching function of reusing Matlab is fitted above-mentioned function (6), is obtaining parameter a1,a2,a3,a4Later,
Finally complete the construction of edge response curve；
5) according to above-mentioned steps 4) the edge response curve function model that is constructed, opposite edges are calculated based on following formulas (7)
Response RER:
Wherein:
D is 0.5 or -0.5；ERx(0.5)、ERxIt (- 0.5) is the opposite edges response in the direction x, ERy(0.5)、ERy(- 0.5) is
The opposite edges response in the direction y.
2. the method according to claim 1 for obtaining the in-orbit focal length of remote sensor based on remote sensing images, which is characterized in that described
The realization of step 3 the following steps are included:
1) the resulting opposite edges response of step 2 is normalized using following formula:
Wherein, xijIndicate the opposite edges response of the jth width emulating image of the i-th width target image, i=1,2 ..., P, j=
1,2,...,Q；
2) correlation between opposite edges response and focal length is analyzed using SPSS software, and is removed and is done according to result
Disturb the data of modeling；
3) data after rejecting are modeled, using opposite edges response focal length each degradation level, that is, focal length grade
On average value as the final data for participating in curve matching, using the opposite edges response after average as variable, with focal length
It carries out curve fitting, obtain a variety of possible function models using the curve estimation of SPSS software itself and its intends as dependent variable
Scale evaluation Value Data is closed, select the wherein maximum function model of fitting degree assessed value loud as the opposite edges finally established
The mathematical model with focal length should be worth.
3. the method according to claim 1 for obtaining the in-orbit focal length of remote sensor based on remote sensing images, which is characterized in that described
The realization of step 4 the following steps are included:
According to the mathematical model that abovementioned steps 3 obtain, when obtaining one group of remote sensing images, by the relative edge for calculating remote sensing images
Edge response, and as input, aforementioned mathematical model is inputted, is exported according to model, obtains the focal length of remote sensor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410520730.4A CN105527072B (en) | 2014-09-30 | 2014-09-30 | A method of the in-orbit focal length of remote sensor is obtained based on remote sensing images |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410520730.4A CN105527072B (en) | 2014-09-30 | 2014-09-30 | A method of the in-orbit focal length of remote sensor is obtained based on remote sensing images |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105527072A CN105527072A (en) | 2016-04-27 |
| CN105527072B true CN105527072B (en) | 2018-12-25 |
Family
ID=55769439
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410520730.4A Active CN105527072B (en) | 2014-09-30 | 2014-09-30 | A method of the in-orbit focal length of remote sensor is obtained based on remote sensing images |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105527072B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107509022B (en) * | 2017-07-21 | 2018-12-21 | 北京空间飞行器总体设计部 | A kind of with task is leading quiet rail Optical remote satellite operating mode implementation method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7038191B2 (en) * | 2003-03-13 | 2006-05-02 | The Boeing Company | Remote sensing apparatus and method |
| CN1799122A (en) * | 2003-05-29 | 2006-07-05 | 松下电器产业株式会社 | Method of measuring focus deviation in pattern exposure and pattern exposure method |
| CN101852623A (en) * | 2010-06-02 | 2010-10-06 | 中国资源卫星应用中心 | On-track calibration method for internal element of satellite optical remote sensing camera |
| CN103115627A (en) * | 2013-01-21 | 2013-05-22 | 武汉大学 | Multi-track combined on-track geometrical checking method of remote sensing satellite linear array sensor |
-
2014
- 2014-09-30 CN CN201410520730.4A patent/CN105527072B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7038191B2 (en) * | 2003-03-13 | 2006-05-02 | The Boeing Company | Remote sensing apparatus and method |
| CN1799122A (en) * | 2003-05-29 | 2006-07-05 | 松下电器产业株式会社 | Method of measuring focus deviation in pattern exposure and pattern exposure method |
| CN101852623A (en) * | 2010-06-02 | 2010-10-06 | 中国资源卫星应用中心 | On-track calibration method for internal element of satellite optical remote sensing camera |
| CN103115627A (en) * | 2013-01-21 | 2013-05-22 | 武汉大学 | Multi-track combined on-track geometrical checking method of remote sensing satellite linear array sensor |
Non-Patent Citations (1)
| Title |
|---|
| 光学遥感卫星在轨参数与图像质量参数关系研究;第7期;《中国优秀硕士学位论文全文数据库信息科技辑》;20140715(第7期);第5、16-20、36-37、41-56、64-65页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105527072A (en) | 2016-04-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110046673B (en) | No-reference tone mapping image quality evaluation method based on multi-feature fusion | |
| CN104658002B (en) | Non-reference image objective quality evaluation method | |
| CN100585634C (en) | A kind of relative radiation correction method of star-load TDICCD camera | |
| CN108401154B (en) | Image exposure degree non-reference quality evaluation method | |
| KR20190075707A (en) | Method for sorting products using deep learning | |
| CN108827181B (en) | Vision-based plate surface detection method | |
| CN102682446A (en) | Method and apparatus for generating a dense depth map using an adaptive joint bilateral filter | |
| CN111127435B (en) | No-reference image quality evaluation method based on double-current convolution neural network | |
| CN108681992A (en) | The image interpolation algorithm of laser facula is measured for detector array method | |
| CN104036493B (en) | No-reference image quality evaluation method based on multifractal spectrum | |
| CN103226819A (en) | Segmental counting-based relative radiation correction method | |
| CN114066857A (en) | Infrared image quality evaluation method and device, electronic equipment and readable storage medium | |
| CN111507426A (en) | No-reference image quality grading evaluation method and device based on visual fusion characteristics | |
| US20140254931A1 (en) | Image Recovery Method | |
| CN103051922B (en) | Establishment method of evaluation criterion parameters and method for evaluating image quality of display screen | |
| Panetta et al. | Logarithmic Edge Detection with Applications. | |
| CN109788275A (en) | Naturality, structure and binocular asymmetry are without reference stereo image quality evaluation method | |
| Pan et al. | No-reference assessment on haze for remote-sensing images | |
| CN113344475A (en) | Transformer bushing defect identification method and system based on sequence modal decomposition | |
| CN114926407A (en) | Steel surface defect detection system based on deep learning | |
| Mahajan et al. | Perceptual quality evaluation of hazy natural images | |
| CN109900363A (en) | A kind of object infrared measurement of temperature method and apparatus based on contours extract | |
| CN111641822B (en) | Method for evaluating quality of repositioning stereo image | |
| CN110567967B (en) | Display panel detection method, system, terminal device and computer readable medium | |
| CN112070762A (en) | Mura defect detection method and device for liquid crystal panel, storage medium and terminal |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |