CN101604017B - Method for realizing remote sensing image simulation under given external orientation element - Google Patents

Abstract

The invention provides a method for realizing remote sensing image simulation under a given external orientation element. The method comprises the following steps: (1) inputting initial data; (2) calculating coordinates of a simulation image element in an image space coordinate system; (3) calculating a spin matrix from the image space coordinate system to an object space coordinate system, and transforming the coordinates of the simulation image element of the step (2) from the image space coordinate system to the object space coordinate system; (4) solving an undetermined proportionality coefficient in a coordinate equation; (5) solving coordinates of a center point of a ground imaging area corresponding to the simulation image element; (6) solving coordinates of four vertexes of the ground imaging area corresponding to the simulation image element; (7) determining an original surficial image element contained in the simulation image element, and calculating a solid angle of the original surficial image element in an instantaneous field of view of the simulation image element; and (8) transforming spatial resolution, and generating a remote sensing simulated image under the external orientation element of a given sensor. The method solves the problem of realizing high precision simulation of the remote sensing image by the external orientation element and a viewing angle under the position and posture of the given sensor.

Description

A kind of method that realizes remote sensing image simulation under the given elements of exterior orientation

Technical field

The present invention relates to a kind of method that realizes remote sensing image simulation under the given elements of exterior orientation, have important value aspect remote sensing images geometry correction and the remotely sensed image emulation, belong to remotely sensed image analogue technique field.

Background technology

The research of remotely sensed image modeling theory and implementation method has important use and is worth in fields such as the prediction of remote sensing task, imaging system design, image quality evaluation and image processing algorithm checkings.Early stage remotely sensed image is simulated mainly based on physical simulation, the development of Along with computer technology, and the remotely sensed image analogue technique has also had new progress, numerous computer simulation emulations technology occurred.

Sensing station and attitude change in time, and this variation has changed the geometric relationship between the sun-ground-sensor three, not only cause ground object the moving of image space on image, cause the geometry deformation of image; And influence the spoke brightness that sensor receives.Less consideration sensing station of existing remote sensing analogue technique and attitude (being elements of exterior orientation) are to the influence of remotely sensed image process.

The minority existing technologies utilizes sensor elements of exterior orientation and field angle or other parameter to set up the geometric relationship between analog image cell coordinate and the ground space coordinate; Calculate corresponding ground point by the analog image cell coordinate, and will use result that numerical value such as this ground point reflectivity carries out remote sensing image simulation as the analog result of treating the analog image pixel.In fact, therefore the corresponding original input picture lastblock of pixel of analog image zone is answered in the surface imaging zone, base area a plurality of topocentric analog results to carry out weighted mean and is generated the analog result of treating the analog image pixel, promptly carries out the spatial resolution conversion.In carrying out the conversion of resolution process, often use average or the distance weighted average and area weighted-average method of equal proportion.Actual conditions are; The analog image pixel is corresponding according to a plurality of pixels in sensor pixels instantaneous field of view angle and the original input picture in ground; The solid angle of original input picture pixel in sensor pixels instantaneous field of view is big more, and this pixel is just big more to the contribution of analog image pixel.Therefore, use average or the distance weighted average and area weighted mean of equal proportion to bring error to analog result.

Summary of the invention

The objective of the invention is to; A kind of method that realizes remote sensing image simulation under the given elements of exterior orientation is provided; To overcome the deficiency of prior art; Set up the geometric relationship between analog image cell coordinate and the corresponding volume coordinate, carry out the spatial resolution conversion, realize remote sensing image simulation under the given elements of exterior orientation according to the solid angle of input picture pixel in sensor pixels instantaneous field of view.

Technical solution of the present invention is: at first utilize sensor elements of exterior orientation and field angle to set up the geometric relationship between analog image cell coordinate and the ground space coordinate; Combine to solve center point coordinate and four apex coordinates of analog image pixel corresponding ground imaging region then with digital elevation model; Last according to the solid angle of original ground surface image picture elements in analog image pixel instantaneous field of view; Carry out the spatial resolution conversion, generate remote sensing analog image under the given elements of exterior orientation.Concrete steps are following:

(1) the input raw data comprises digital elevation model, original ground surface image and sensor elements of exterior orientation and field angle;

(2) under the sensor field of view angle of setting in the calculation procedure (1), analog image pixel coordinate in image space coordinate system;

(3) under the sensor elements of exterior orientation of setting in the calculation procedure (1), image space coordinate system is to the rotation matrix of object space coordinate system, and the middle analog image cell coordinate of step (2) is transformed into object space coordinate system from image space coordinate system;

(4) by the digital elevation model and the middle analog image pixel equation in coordinates in object space coordinate system of step (3) of step (1) input, find the solution scale-up factor undetermined in the equation in coordinates;

(5) bring the scale-up factor that obtains in the step (4) in the step (3) analog image pixel equation in coordinates in object space coordinate system, obtain the center point coordinate of analog image pixel corresponding ground imaging region;

(6) revise analog image pixel coordinate in image space coordinate system in the step (2),, obtain four apex coordinates of analog image pixel corresponding ground imaging region according to step (3), step (4) and step (5);

(7) ask for the original ground surface image picture elements of input in the step (1) that comprises in the imaging region of ground, and calculate the solid angle of original ground surface image picture elements in analog image pixel instantaneous field of view;

(8) according to the solid angle of original ground surface image picture elements in analog image pixel instantaneous field of view, carry out the spatial resolution conversion, generate remote sensing analog image under the given sensor elements of exterior orientation.

Wherein, the coordinate in image space coordinate system of the analog image pixel described in the step (2) is:

$(\tan ((x-\frac{M}{2})\frac{{\mathrm{FOV}}_c}{M}+\mathrm{\Φ})$ $\tan ((y-\frac{N}{2})\frac{{\mathrm{FOV}}_d}{N}+\mathrm{\θ}),-1)$

Wherein, x and y are the row of analog image pixel in analog image number and row number, and M and N are the line number and the columns of analog image, FOV _dFor sensor along field angle on the rail direction, FOV _cFor sensor is being handed over field angle on the rail direction, Φ, θ are respectively along the rail direction and hand over field angle residual bias on the rail direction.

Wherein, the coordinate in object space coordinate system of the analog image pixel described in the step (3) is:

$\left(\begin{array}{c}X\\ Y\\ Z\end{array}\right)=\mathrm{\λR}\left(\begin{array}{c}\tan ((x-\frac{M}{2}){\mathrm{FOV}}_c/M+\mathrm{\Φ})\\ \tan ((y-\frac{N}{2}){\mathrm{FOV}}_d/N+\mathrm{\θ})\\ -1\end{array}\right)+\left(\begin{array}{c}\mathrm{Xs}\\ \mathrm{Ys}\\ \mathrm{Zs}\end{array}\right)$

Wherein, x and y are the row of analog image pixel in analog image number and row number, and M and N are the line number and the columns of analog image, FOV _dFor sensor along field angle on the rail direction, FOV _cFor sensor is being handed over field angle on the rail direction, Φ, θ are respectively along the rail direction and hand over field angle residual bias on the rail direction, (X; Y Z) is analog image pixel coordinate in object space coordinate system, (Xs; Ys Zs) is center sensor coordinate in object space coordinate system, i.e. sensor outer orientation line element; λ is a scale-up factor, image space coordinate system to the rotation matrix R of object space coordinate system by sensor foreign side parallactic angle element

ω and κ decision:

Wherein, the analog image pixel described in the step (4) in the object space coordinate system equation in coordinates scale-up factor undetermined find the solution according to:

At first found the solution the initial value λ ini of scale-up factor λ by elevation maximal value Zm in the input digit elevation model in the step (1), λ ini=Zs-Zm is an initial value with λ ini then; According to progressively increasing λ for fixed step size t (t＞0); Work as i=0,1,2; 3...... the time corresponding scale-up factor λ i=λ ini+it, calculate object space coordinate (X this moment _{λ i}, Y _{λ i}, Z _{λ i}), obtain (X by digital elevation model _{λ i}, Y _{λ i}) the corresponding actual height value Zi in ground, relatively Zi and Z _λ, if Z _λ-Zi is less than given threshold value, and then this moment, λ i was the scale-up factor of asking; If Z _λ-Zi then continues to increase λ i greater than given threshold value, calculates (X _{λ i}, Y _{λ i}, Z _{λ i}) and Zi, until Z _λ-Zi finally obtains scale-up factor λ undetermined less than given threshold value.

Wherein, described in the step (6) find the solution four apex coordinates of analog image pixel corresponding ground imaging region the time, analog image pixel coordinate in image space coordinate system is modified to:

$(\tan ((x-\frac{M}{2}){\mathrm{FOV}}_c/M+a+\mathrm{\Φ}),\tan ((y-\frac{N}{2}){\mathrm{FOV}}_d/N+b+\mathrm{\θ}),-1)$

Wherein, a value is respectively FOV _c/ 2M, FOV _c/ 2M ,-FOV _c/ 2M and-FOV _c/ 2M; Corresponding b value with a is respectively FOV _d/ 2N ,-FOV _d/ 2N ,-FOV _d/ 2N and FOV _d/ 2N.

Wherein, the original ground surface image picture elements described in the step (7) the calculating at analog image pixel instantaneous field of view neutral body angle according to:

${\mathrm{\Ω}}_i=\frac{A_i\cos {\mathrm{\θ}}_i}{r_i^2}$

Wherein, Ω _iBe the solid angle of original ground surface image picture elements i in treating analog image pixel instantaneous field of view, A _iBe the area of original ground surface image picture elements i in treating analog image pixel instantaneous field of view, θ _iFor original ground surface image picture elements i to the angle between center sensor line and the ground normal, r _iBe the distance of original ground surface image picture elements i to center sensor.

Wherein, described in the step (8) according to the solid angle of original ground surface image picture elements in analog image pixel instantaneous field of view, carry out spatial resolution conversion:

$L=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\frac{{\mathrm{\Ω}}_i}{\mathrm{\Ω}}{l}_i$

Wherein, L is the brightness value of analog image pixel, and the number of the original ground surface image picture elements that i=1,2,3...N comprise for this analog image pixel, Ω are the instantaneous field of view angle of this analog image pixel, Ω _iBe the solid angle of original ground surface image picture elements i in treating analog image pixel instantaneous field of view, l _iThe brightness value of the original ground surface image picture elements i that comprises for the analog image pixel.

A kind of method that realizes remote sensing image simulation under the given elements of exterior orientation of the present invention, advantage compared with prior art is:

(1) the present invention considers that the sensing station attitude changes the influence to the remotely sensed image process; Utilize sensor elements of exterior orientation and field angle to set up the geometric relationship between analog image cell coordinate and the ground space coordinate, can realize the accurate simulation of remote sensing images under any given sensor elements of exterior orientation.

(2) the present invention accurately finds the solution four apex coordinates of analog image pixel corresponding ground imaging region; And by a plurality of ground points in the imaging region of ground; Carry out the spatial resolution conversion according to the solid angle weighted mean of original ground surface image picture elements in analog image pixel instantaneous field of view; Synthetic wait to simulate the pixel brightness value, overcome and used that equal proportion is average or distance weighted on average carries out the error that spatial resolution is changed with the area weighted mean.

Description of drawings

Fig. 1 realizes the process flow diagram of remote sensing image simulation method under the given elements of exterior orientation for the present invention.

Embodiment

The simulation process of driving remote sensing images under the regional given sensor elements of exterior orientation of dragon with Tibet is example, and as shown in Figure 1, practical implementation method of the present invention is following:

(1) the input raw data comprises digital elevation model, original ground surface image and sensor elements of exterior orientation and field angle.

Use Tibet to drive digital elevation model and the face of land spoke brightness data of imperial regional ground resolution as 30m.With the face of land upper left angle point of digital elevation model data is the center, and 30m is a unit length, and due east Zheng Bei is respectively x axle and y axle, and the elevation direction is the z axle; Set up object space coordinate system, the sensor elements of exterior orientation is (1678,500; 23500) and (0 °, 0 °, 0 °); Sensor is being respectively 1 ° and 0.8 ° along field angle on rail direction and the friendship rail direction, and analog image line number and columns are 250 and 200, all is made as 0 along field angle residual bias on rail direction and the friendship rail direction.

(2) under the sensor field of view angle of setting in the calculation procedure (1), analog image pixel coordinate in image space coordinate system.

Under the sensor field of view angle of in step (1), setting, the analog image pixel (x, y) coordinate is in image space coordinate system:

(tan(0.004x-0.5)，tan(0.004y-0.4)，-1)

(3) under the sensor elements of exterior orientation of setting in the calculation procedure (1), image space coordinate system is to the rotation matrix of object space coordinate system, and the middle analog image cell coordinate of step (2) is transformed into object space coordinate system from image space coordinate system.

Under the sensor elements of exterior orientation of in step (1), setting, rotation matrix R is a unit matrix, the analog image pixel (x, y) in object space coordinate system coordinate (X, Y Z) are:

$\left(\begin{array}{c}X\\ Y\\ Z\end{array}\right)=\mathrm{\λ}\left(\begin{array}{c}\tan (0.004x-0.5)\\ \tan (0.004y-0.4)\\ -1\end{array}\right)+\left(\begin{array}{c}1678\\ 500\\ 23500\end{array}\right)$

Wherein, λ is a scale-up factor undetermined.

(4) by the digital elevation model and the middle analog image pixel equation in coordinates in object space coordinate system of step (3) of step (1) input, find the solution scale-up factor undetermined in the equation in coordinates.

The analog image pixel in object space coordinate system in the equation in coordinates scale-up factor find the solution according to:

At first found the solution the initial value λ ini of scale-up factor λ by elevation maximal value Zm in the input digit elevation model in the step (1), then λ ini=Zs-Zm is an initial value with λ ini then; According to giving fixed step size t (t＞0) progressively scaling up coefficient; Work as i=0,1,2; 3...... the time corresponding scale-up factor λ i=λ ini+it, calculate object space coordinate (X this moment _{λ i}, Y _{λ i}, Z _{λ i}), can access (X by digital elevation model _{λ i}, Y _{λ i}) corresponding actual ground height value Zi, relatively Zi and Z _λ, if Z _λ-Zi is less than given threshold value, and then this moment, λ i was the scale-up factor of asking; If Z _λ-Zi then continues to increase λ i greater than given threshold value, calculates (X _{λ i}, Y _{λ i}, Z _{λ i}) and Zi, until Z _λ-Zi finally obtains scale-up factor λ undetermined less than given threshold value.

(5) bring the scale-up factor that obtains in the step (4) in the step (3) analog image pixel equation in coordinates in object space coordinate system, obtain the center point coordinate of analog image pixel corresponding ground imaging region.

(6) revise analog image pixel coordinate in image space coordinate system in the step (2),, obtain four apex coordinates of analog image pixel corresponding ground imaging region according to step (3), step (4) and step (5).

Analog image pixel coordinate in image space coordinate system is modified to:

(tan(0.004x-0.5+a)，tan(0.004y-0.4+b)，-1)

Wherein, a value is respectively 0.002,0.002 ,-0.002 and-0.002; Corresponding b value with a is respectively 0.002 ,-0.002 ,-0.002 and 0.002.

(7) ask for the original ground surface image picture elements of input in the step (1) that comprises in the imaging region of ground, and calculate the solid angle of original ground surface image picture elements in analog image pixel instantaneous field of view.

The original ground surface image picture elements the calculating of treating analog image pixel instantaneous field of view neutral body angle according to:

${\mathrm{\Ω}}_i=\frac{A_i\cos {\mathrm{\θ}}_i}{r_i^2}$

Wherein, Ω _iBe the solid angle of original ground surface image picture elements i in treating analog image pixel instantaneous field of view, A _iBe original ground surface image picture elements i area in treating the analog image pixel, θ _iFor original ground surface image picture elements i to the angle between center sensor line and the ground normal, r _iBe the distance of original ground surface image picture elements i to center sensor.

(8) according to the solid angle of original ground surface degree image picture elements in analog image pixel instantaneous field of view, carry out the spatial resolution conversion, generate remote sensing analog image under the given sensor elements of exterior orientation.

According to the solid angle of original ground surface image picture elements in analog image pixel instantaneous field of view, carry out the spatial resolution conversion:

$L=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\frac{{\mathrm{\Ω}}_i}{\mathrm{\Ω}}{l}_i$

Wherein, L is the brightness value of analog image pixel, and the number of the original ground surface image picture elements that i=1,2,3...N comprise for this analog image pixel, Ω are the instantaneous field of view angle of this analog image pixel, Ω _iBe the solid angle of original ground surface image picture elements i in treating analog image pixel instantaneous field of view, l _iThe brightness value of the original ground surface image picture elements i that comprises for the analog image pixel.

Claims (4)

1. method that realizes remote sensing image simulation under the given elements of exterior orientation is characterized in that may further comprise the steps:

(1) the input raw data comprises digital elevation model, original ground surface image and sensor elements of exterior orientation and field angle;

(2) under the sensor field of view angle of setting in the calculation procedure (1), analog image pixel coordinate in image space coordinate system;

(3) under the sensor elements of exterior orientation of setting in the calculation procedure (1), image space coordinate system is to the rotation matrix of object space coordinate system, and the middle analog image cell coordinate of step (2) is transformed into object space coordinate system from image space coordinate system;

(4) by the digital elevation model and the middle analog image pixel equation in coordinates in object space coordinate system of step (3) of step (1) input, find the solution scale-up factor undetermined in the equation in coordinates;

(5) bring the scale-up factor that obtains in the step (4) in the step (3) analog image pixel equation in coordinates in object space coordinate system, obtain the center point coordinate of analog image pixel corresponding ground imaging region;

(6) revise analog image pixel coordinate in image space coordinate system in the step (2), according to step

(3), step (4) and step (5), obtain four apex coordinates of analog image pixel corresponding ground imaging region;

(7) ask for the original ground surface image picture elements of input in the step (1) that comprises in the imaging region of ground, and calculate the solid angle of original ground surface image picture elements in analog image pixel instantaneous field of view;

(8) according to the solid angle of original ground surface image picture elements in analog image pixel instantaneous field of view, carry out the spatial resolution conversion, generate remote sensing analog image under the given sensor elements of exterior orientation.

2. a kind of method that realizes remote sensing image simulation under the given elements of exterior orientation according to claim 1 is characterized in that: analog image pixel coordinate in image space coordinate system is in the described step (2):

$(\tan ((x-\frac{M}{2})\frac{{\mathrm{FOV}}_c}{M}+\mathrm{\Φ}),\tan ((y-\frac{N}{2})\frac{{\mathrm{FOV}}_d}{N}+\mathrm{\θ}),-1)$

Wherein, x and y are the row of analog image pixel in analog image number and row number, and M and N are the line number and the columns of analog image, FOV _dFor sensor along field angle on the rail direction, FOV _cFor sensor is being handed over field angle on the rail direction, Φ, θ are respectively along the rail direction and hand over field angle residual bias on the rail direction.

3. a kind of method that realizes remote sensing image simulation under the given elements of exterior orientation according to claim 1; It is characterized in that: when finding the solution four apex coordinates of analog image pixel corresponding ground imaging region in the described step (6), analog image pixel coordinate in image space coordinate system is modified to:

$(\tan ((x-\frac{M}{2}){\mathrm{FOV}}_c/M+a+\mathrm{\Φ}),\tan ((y-\frac{N}{2}){\mathrm{FOV}}_d/N+b+\mathrm{\θ}),-1)$

Wherein, a value is respectively FOV _c/ 2M, FOV _c/ 2M ,-FOV _c/ 2M and-FOV _c/ 2M; Corresponding b value with a is respectively FOV _d/ 2N ,-FOV _d/ 2N ,-FOV _d/ 2N and FOV _d/ 2N;

Wherein, x and y are the row of analog image pixel in analog image number and row number, and M and N are the line number and the columns of analog image, FOV _dFor sensor along field angle on the rail direction, FOV _cFor sensor is being handed over field angle on the rail direction, Φ, θ are respectively along the rail direction and hand over field angle residual bias on the rail direction.

4. a kind of method that realizes remote sensing image simulation under the given elements of exterior orientation according to claim 1; It is characterized in that: according to the solid angle of original ground surface image picture elements in analog image pixel instantaneous field of view, carry out the spatial resolution conversion in the described step (8):

$L=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\frac{{\mathrm{\Ω}}_i}{\mathrm{\Ω}}{l}_i$

Wherein, L is the brightness value of analog image pixel, and the number of the original ground surface image picture elements that i=1,2,3...N comprise for this analog image pixel, Ω are the instantaneous field of view angle of this analog image pixel, Ω _iBe the solid angle of original ground surface image picture elements i in treating analog image pixel instantaneous field of view, l _iThe brightness value of the original ground surface image picture elements i that comprises for the analog image pixel.

Images