CN1308657C - Method for measuring formation of seamless space stereomodel - Google Patents

Abstract

The present invention discloses a method for forming a measurable seamless space stereomodel. The method for forming a measurable seamless space stereomodel by using the projection function comprises the following steps: orderly numbering aerial photographs in the direction of a flight strip; making orthophotos by single number films 1, 3, 5, etc., by using digital ground models; making supplementary films of the orthophotos by double number films 2, 4, 6, etc., by using the digital ground models; embedding the digital orthophotos made of odd number films into a seamless area according to the ground coordinates; simultaneously, recording the area range of the digital orthophotos after each digital orthophoto is embedded; the measurable seamless stereomodel is constructed by a point-to-point method. The stereomodel formed with the method of the present invention breaks through the limitation from the photo measuring neutral body image to the ranges. In the method, the seamless space stereomodel can be conveniently constructed and used without photographing and measuring professional knowledge, the three-dimensional terrain landscape model during photographing can be visually restored and reproduced, the seamless stereo roaming and browsing can be carried out, and the three-dimensional measurement and information collection on the stereomodel can be carried out. The method can be used in other relative industries, such as survey, geology, forestry, electric power, city planning, highway design, railway design, etc.

Description

Method for measuring formation of seamless space stereomodel

Technical field

The invention belongs to the Surveying Science and Technology field, relate to a kind of method of utilizing aviation image and digital terrain model (DEM) to generate a kind of convenient measured seamless space stereoscopic model that makes up and use.

Background technology

Our nature of life is a three dimensions, and most objects that we saw all are three-dimensional, and what people were accustomed to is to measure an object with three parameters of length.Three-dimensional information major part is that the eyes by the people obtain.As the saying goes " it is better to see once than hear a hundred times ", this has vividly described the vital role of eyes very much, and human eye is that the people observes, the most important organ in the perception world thus.In the face of the gorgeous charming scenery of the Nature, people wish to seek a kind of new 3 D displaying method when praise is sighed with feeling, realistically, three-dimensionally reappear natural landscape.Because traditional display packing is the two-dimensional projection of three-dimensional scenic mostly, wherein most of three-dimensional depth information has been lost, and people can only by virtue of experience judge the degree of depth level of object in the scene, and this is very false.People study the principle of stereoscopic vision of human eye, thereby make artificial realization stereo display become possibility by the further investigation to human eye.Stereoscopic vision is extremely important to the people, does not have the world of three-dimensional level to imagine.The people observes by eyes and has produced stereoscopic visual effect.Normal people recognizes the three dimensions object with eyes.People are when watching a certain three-dimensional body, people's eyes are observed from the differentiated slightly angle of the right and left, therefore observed object formed optical image on the retina of people's right and left eyes is slightly variant, and the usually said binocular parallax (Binocular Parallax) of this species diversity people also is the physiology parallax.The generation of parallax mainly is because between people's the right and left eyes certain distance is arranged, and adult eyes distance is about 65mm.The generation of vision has important effect to the formation of stereoscopic vision.Optical image passes simultaneously to brain optic nerve maincenter on the right and left eyes retina, and the left and right sides optical image of parallax is arranged, through the processing and the fusion of optic nerve, the people just can experience seen the three-dimensional level of object, artificial stereoscopic vision that Here it is.

In photogrammetric field, artificial stereoscopic vision has been brought into use for a long time.In latter stage in 18th century, the general Frisch (C.Pulfrich) of Jena, Germany Zeiss factory has proposed the principle of stereopsis, and has made stereocomparator in 1901, is called as " father of stereophotogrammetry " in Germany.Be photogrammetric basis of establishing stereo measurement simultaneously.After this double projection mapping, stereoscopic plotter, planicomp all are based on this principle up to the totally digitilized digital mapping that develops into present digital photogrammetric work station.

The history of existing centuries of photogrammetric development, analog photogrammetry, analytical photogrammetry and three developing stage of digital photogrammetry have been experienced, space multistory model when recovering photography by the method for simulation, parsing or numeral, atural object and landforms by stereoscopic model true reappearance area being shot, on model, carry out stereo measurement, gather some basic geography information, this is a photogrammetric principal character.Because the foundation of stereoscopic model will be through complicated interior orientation, as to orientation and absolute orientation Photogrammetric Processing process, the professional knowledge that must possess photogrammetric specialty, therefore, this spatial model is to be set up and use by photogrammetric worker basically, is photogrammetric worker's a patent.

This traditional by about the stereoscopic model that constitutes of two images be photogrammetric basis and core, since its scope only limit to two photos overlapping very among a small circle in, and set up the process complexity, therefore, the media of gathering as just photogrammetric worker's a kind of geodata, after finishing data collection task, just be abandoned, again be not used.Because all information of topographical surface when containing photography in the stereoscopic model, and be very limited, if, be one and make us very regret feelings stereoscopic model losing in vain through the information that is remained after the data acquisition.If we can construct a large-scale and convenient space multistory model that makes up; just can make the sand smeller; the forest scholar; the planning scholar; slip-stick artist and other industry are used the people of airphoto; under the prerequisite that does not need photogrammetric complicated knowledge background; make up and the usage space stereoscopic model with simple method; to be planning efficiently; the exploitation and protect national resource and make great contribution; simultaneously; the layman also can be sitting in family by spatial model and browse dimensional topography landscape model true to nature, can increase the value of aeroplane photography achievement greatly.

Summary of the invention

Problem to be solved by this invention is: a kind of generation method of the convenient scalable seamless space stereoscopic model that makes up and use on a large scale is provided, the stereoscopic model that this method is constructed has been broken through the limitation of photogrammetric neutral body picture to scope, do not need photogrammetric professional knowledge conveniently to make up and to use, dimensional topography landscape model in the time of can recovering and reproduce photography really, can carry out seamless solid and roam and browse, and can on model, carry out three-dimensional measuring and information acquisition.

Technical scheme provided by the invention is: a kind of method for measuring formation of seamless space stereomodel, and its step is as follows:

(1) according to direction airphoto is carried out serial number along air strips;

(2) utilize digital terrain model, make orthography with odd numbers sheet 1,3,5... sheet, concrete making step is as follows:

I), obtain four angular coordinate (X of square grid to the netting index word elevation model according to orthogonal projection with the square grid of certain intervals on the XY plane _i, Y _i, Z _i)

Ii) by collinearity condition equation

$x-x_0=-f\frac{a_1(X-X_S)+b_1(Y-Y_S)+c_1(Z-Z_S)}{a_3(X-X_S)+b_3(Y-Y_S)+c_3(Z-Z_S)}$

$y-y_0=-f\frac{a_2(X-X_S)+b_2(Y-Y_S)+c_2(Z-Z_S)}{a_3(X-X_S)+b_3(Y-Y_S)+c_3(Z-Z_S)}$

In the formula: x, y ,-f are picture point image space coordinate;

F is the focal length of aerophoto;

x ₀, y ₀Be the principal point coordinate;

X, Y, Z are topocentric object space coordinate;

X _S, Y _S, Z _SFor taking the photograph the object space coordinate of website;

a _i, b _i, c _i(i=1,2,3) are that three elements of exterior orientation of photo are formed nine direction cosine;

With step I) four angular coordinates intersecting with digital terrain model after the projection transform to photographic coordinate system according to collinearity condition equation and obtain corresponding picpointed coordinate (x _i, y _i);

Iii) calculate corresponding scanning coordinate by picpointed coordinate by the interior orientation transformation parameter

$\left[\begin{array}{c}I\\ J\end{array}\right]=\left[\begin{array}{cc}{m}_1& m_2\\ n_1& n_2\end{array}\right]\left[\begin{array}{c}x-x_0\\ y-y_0\end{array}\right]-\left[\begin{array}{c}{I}_0\\ J_0\end{array}\right]$

M in the formula ₁, m ₂, n ₁, n ₂, I ₀, J ₀Be the interior orientation parameter;

Iv) to the scanning coordinate of the pixel in each piece according to bilinearity polynomial interpolation correspondence

I＝a ₀+a ₁X+a ₂Y+a ₃XY

J＝b ₀+b ₁X+b ₂Y+b ₃XY

I in the formula, J are scanning coordinate, X, and Y is ground coordinate, a _i, b _iBe the bilinear transformation coefficient

V) gray scale interpolation

Four angle points of each bin are obtained its photo coordinate according to collinearity equation, be translated into scanning coordinate by the interior orientation parameter then, calculate its scanning coordinate for each piece inside according to the bilinear transformation pointwise, adopt the gray-scale value of each pixel of gray scale interpolating method interpolation again;

Vi) gray scale assignment

Each gray-scale value after the interpolation is composed one by one to each pixel after correcting;

Each point is promptly formed digital orthoimage after all handling;

(3) utilize digital terrain model, with even numbers sheet 2,4,6 ... make the supplementary plate of orthography, the process of making and the process of orthography are similar, step I just) projection function of introducing parallax method with the varied angle projection carries out projection and orthogonal projection when replacing making orthography, and the projection function of described varied angle projection introducing parallax method is $P_i=\frac{{\mathrm{BZ}}_i}{H-Z_i},$ P in the formula _iBe the parallax of each point, Z _iBe the height value of each point, B is the photographic base of stereogram, the flying height when H is photography; Concrete making step is as follows:

By the square grid on the XY plane, get projecting direction according to the projection function of varied angle projection introducing parallax method and be parallel to XZ plane introducing parallax, according to the computing formula of the intersection point on varied angle projection ray and digital terrain model surface, with four angle points of graticule mesh project to digital elevation model obtain projection line and digital terrain model intersection point (X ' _iY ' _iZ ' _i), following calculation procedure goes on foot to vi with the ii that makes orthography;

(4) digital orthoimage that the odd number sheet is made is set into a seamless zone according to ground coordinate with it, writes down the regional extent of the digital orthoimage of each digital orthography after inlaying simultaneously;

(5) the scalable seamless stereoscopic model of point-to-point structure.

Ultimate principle of the present invention is based on the principle of stereoscopic vision of human eye, based on digital orthoimage, utilize original airphoto and digital terrain model (DEM), generating digital orthoimage by the odd number sheet, utilize the even number sheet to introduce parallax and generate a stereo normal incidence image supplementary plate that is complementary with orthography by the parallax function, digital orthoimage and corresponding auxiliary image constitute orthophoto stereo-mate together, after digital orthoimage is carried out jointless mosaic, utilize the construction algorithm of stereoscopic model to form the seamless supreme scalable landscape model of three-dimensional of parallax down on a large scale with supplementary plate.

Use the stereoscopic model that the present invention constructed to break through the limitation of photogrammetric neutral body picture to scope, do not need photogrammetric professional knowledge conveniently to make up and to use, dimensional topography landscape model in the time of can recovering and reproduce photography really, can carry out seamless solid and roam and browse, and can on model, carry out three-dimensional measuring and information acquisition.Can use for other relevant industries such as mapping, geology, forestry, electric power, city planning, highway and Railway Design.

Description of drawings

Fig. 1 is with respect to two disparity maps that point is produced of reference field differing heights on the ground of original when photography;

Fig. 2 is the introducing disparity map identical with original parallax;

Fig. 3 is an orthophotoquad;

Fig. 4 is the orthography supplementary plate;

Fig. 5 is the projection ray and the DEM intersection point calculation figure of varied angle;

Fig. 6 is the photo numbering;

Fig. 7 is a process flow diagram of the present invention.

Embodiment

At first introduce varied angle projection of the present invention below and introduce parallax method.

As shown in Figure 1 and Figure 2, Fig. 1 be on the ground of original when photography with respect to two parallaxes that point is produced of reference field differing heights, Fig. 2 is identical parallax and the α that adopts in order to produce with original photography ₁And α ₂The projection ray of two different angles, the geometric relationship from Fig. 1, Fig. 2 as can be known:

$\mathrm{tg}{\mathrm{\&alpha;}}_1=\frac{\mathrm{<figure-callout\; id="1"\; label="B\; Z"\; filenames="C0214775200105.png"\; state="\{\{state\}\}">B</figure-callout>}}{Z_{}}$ $\mathrm{tg}{\mathrm{\&alpha;}}_2=\frac{B}{Z_2}---\left(1\right)$

B is the photographic base of stereogram in the formula, flying height (as follows) when H is photography, therefore, for the parallax with original when photography is consistent, adopt the varied angle projection to introduce parallax, Fig. 3, Fig. 4 have shown the principle of varied angle introducing parallax, and employing is parallel to the XZ plane and carries out projection, according to different ground elevations, the angle of the projection ray of each point is different.

As the geometric relationship of Fig. 3, Fig. 4 as can be known, varied angle is introduced the projection function of parallax method and is:

$P_i=\frac{{\mathrm{BZ}}_i}{H-Z_i}---\left(2\right)$

P in the formula _iParallax for each point; Z _iHeight value for each point.

By following formula as can be known, the varied angle projection is introduced in the orthophoto stereo-mate of parallax arbitrarily the elevation of any and can be obtained by following formula:

$Z_i=\frac{P_{i}H}{B+P_i}---\left(3\right)$

Because when generating the stereo normal incidence image supplementary plate, must calculate the intersection point on projection function and digital terrain model (DEM) surface, the intersection point on varied angle projection ray and DEM surface (X ' _iY ' _iZ ' _i) as shown in Figure 5, as the geometric relationship among the figure as can be known:

$Y_i^{\&prime;}=Y_i$

$X_i^{\&prime;}=\frac{(X_{i+1}-X_i)(X_i+k{Z}_i)-X_{i}k(Z_{i+1}-Z_i)}{X_{i+1}-k(Z_{i+1}-Z_i)-X_i}---\left(4\right)$

$Z_i^{\&prime;}=Z_i+\frac{(Z_{i+1}-Z_i)(X_i^{\&prime;}-X_i)}{(X_{i+1}-X_i)}$

In the formula $k=\tan \mathrm{\&alpha;}=\frac{B}{Z_i}$

What need any be described is, when landform big rise and fall or DEM graticule mesh comparatively dense, when really finding the solution varied angle projection ray and DEM intersection point, must judge earlier which DEM mesh spacing parallel rays drops on, and could separate the intersection point of asking parallel rays and DEM with top formula then.

Describe varied angle above in detail and introduce the projection function of parallax method, introduce below and utilize this projection function to generate the step (referring to Fig. 7) that can measure the seamless space stereoscopic model:

(1) collects raw data (raw video, DEM and photo parameter); (be followed successively by 1,2,3,4,5,6 according to airphoto being carried out serial number then along the direction of air strips ...), as shown in Figure 6.

(2) orthography making-digital orthoimage-orthography jointless mosaic;

Utilize digital terrain model DEM, with odd numbers sheet 1,3,5 ... sheet is made orthography, and concrete making step is as follows:

I) with the square grid of certain intervals on the XY plane graticule mesh equal and opposite in direction of digital elevation model (general with), according to four angular coordinate (X of orthogonal projection acquisition square grid to the netting index word elevation model _i, Y _i, Z _i)

Ii) by collinearity condition equation

$x-x_0=-f\frac{a_1(X-X_S)+b_1(Y-Y_S)+c_1(Z-Z_S)}{a_3(X-X_S)+b_3(Y-Y_S)+c_3(Z-Z_S)}---\left(5\right)$

$y-y_0=-f\frac{a_2(X-X_S)+b_2(Y-Y_S)+c_2(Z-Z_S)}{a_3(X-X_S)+b_3(Y-Y_S)+c_3(Z-Z_S)}$

In the formula: x, y ,-f are picture point image space coordinate;

F is the focal length of aerophoto;

x ₀, y ₀Be the principal point coordinate;

X, Y, Z are topocentric object space coordinate;

X _S, Y _S, Z _SFor taking the photograph the object space coordinate of website;

a _i, b _i, c _i(i=1,2,3) are that three elements of exterior orientation of photo are formed nine direction cosine;

With step I) four angular coordinates intersecting with DEM after the projection transform to photographic coordinate system according to collinearity condition equation and obtain corresponding picpointed coordinate (x _i, y _i).

Iii) calculate corresponding scanning coordinate by picpointed coordinate by the interior orientation transformation parameter

$\left[\begin{array}{c}I\\ J\end{array}\right]=\left[\begin{array}{cc}{m}_1& m_2\\ n_1& n_2\end{array}\right]\left[\begin{array}{c}x-x_0\\ y-y_0\end{array}\right]-\left[\begin{array}{c}{I}_0\\ J_0\end{array}\right]---\left(6\right)$

M in the formula ₁, m ₂, n ₁, n ₂, I ₀, J ₀Be the interior orientation parameter;

Iv) to the scanning coordinate of the pixel in each piece according to bilinearity polynomial interpolation correspondence

I＝a ₀+a ₁X+a ₂Y+a ₃XY (7)

J＝b ₀+b ₁X+b ₂Y+b ₃XY

I, J are scanning coordinate

X, Y are ground coordinate

a _i, b _iBe the bilinear transformation coefficient

V) gray scale interpolation

Four angle points of each bin are obtained its photo coordinate according to collinearity equation, be translated into scanning coordinate by the interior orientation parameter then, calculate its scanning coordinate for each piece inside according to the bilinear transformation pointwise, adopt the gray-scale value of each pixel of gray scale interpolating method interpolation again.

Vi) gray scale assignment

Each gray-scale value after the interpolation is composed one by one to each pixel after correcting.

Each point is promptly formed digital orthoimage after all handling.

(3) auxiliary making video-supplementary plate image;

Utilize digital terrain model DEM, with even numbers sheet 2,4,6 ... make the supplementary plate of orthography, the process of making and the process of orthography are similar, step I just) carries out projection with projection function and orthogonal projection when replacing making orthography, concrete making step is as follows: by the square grid on the XY plane, get projecting direction according to varied angle projection introducing parallax method and be parallel to XZ plane introducing parallax, according to the computing formula of the intersection point on varied angle projection ray and DEM surface, with four angle points of graticule mesh project to digital elevation model obtain projection line and DEM intersection point (X ' _iY ' _iZ ' _i), following calculation procedure goes on foot to vi with the ii that makes orthography.

(4) digital orthoimage that the odd number sheet is made is set into a seamless zone according to ground coordinate with it, writes down the regional extent of the digital orthoimage of each digital orthography after inlaying simultaneously.

(5) the scalable seamless stereoscopic model of point-to-point structure.

Only there is the parallax on the horizontal direction owing to do not have the parallax on the vertical direction between the supplementary plate of digital orthoimage and orthography, according to the ultimate principle of the structure of stereoscopic vision, can form stereoscopic vision as long as guarantee two width of cloth images that same object is obtained from different positions.Because digital orthoimage and corresponding orthography supplementary plate left sheet and the right sheet during respectively from photography, and engineer's scale unanimity, satisfy the pacing items of constructing stereo vision, there is not simultaneously vertical parallax again, therefore, as long as the some pointwise correspondence on each point on the assurance orthography and the orthography supplementary plate can form solid geometric pattern.So because digital orthoimage is the seamless solid geometric pattern that constructs based on orthography also is seamless.Specific algorithm is as follows:

I) point in the peek word orthography calculates the regional extent that this coordinate drops on that the digital orthography before inlaying according to the ground of this point, finds its photo numbering, can find corresponding orthography coupling sheet then.

It is right ii) to find corresponding same place (be same object on two photos the point of imaging) respectively promptly to constitute a point in the stereoscopic model in orthography coupling sheet according to the ground coordinate of this point.

After iii) pointwise is handled like this, can guarantee that each point in the model all is by right from the solid point of digital orthoimage and digital orthoimage coupling sheet, and then form seamless stereoscopic model.

(6) stereo display of stereoscopic model (complementary color method or stroboscopic method) and measurement

Adopt liquid crystal glasses as stereoscopy equipment (also can adopt alternate manner), utilize the double buffering stereo display mechanism of Open GL respectively each three-dimensional point to be caused display buffer, the left and right sides to sending, promptly can be observed three-dimensional stereo model by liquid crystal glasses.Record behind the parallax of each some the elevation information that the formula that calculates elevation by parallax by the front can calculate every bit by the 3 D stereo mark, simultaneously, its planimetric position can be obtained by digital orthoimage.

Claims (1)

1. method for measuring formation of seamless space stereomodel, its step is as follows:

(1) according to direction airphoto is carried out serial number along air strips;

(2) utilize digital terrain model, make orthography with odd numbers sheet 1,3,5... sheet, concrete making step is as follows:

I), obtain four angular coordinate (X of square grid to the netting index word elevation model according to orthogonal projection with the square grid of certain intervals on the XY plane _i, Y _i, Z _i)

Ii) by collinearity condition equation

$x-x_0=-f\frac{a_1(X-X_S)+b_1(Y-Y_S)+c_1(Z-Z_S)}{a_3(X-X_S)+b_3(Y-Y_S)+c_3(Z-Z_S)}$

$y-y_0=-f\frac{a_2(X-X_S)+b_2(Y-Y_S)+c_2(Z-Z_S)}{a_3(X-X_S)+b_3(Y-Y_S)+c_3(Z-Z_S)}$

In the formula: x, y ,-f are picture point image space coordinate;

F is the focal length of aerophoto;

x ₀, y ₀Be the principal point coordinate;

X, Y, Z are topocentric object space coordinate;

X _S, Y _S, Z _SFor taking the photograph the object space coordinate of website;

a _i, b _i, c _i(i=1,2,3) are that three elements of exterior orientation of photo are formed nine direction cosine;

With step I) four angular coordinates intersecting with digital terrain model after the projection transform to photographic coordinate system according to collinearity condition equation and obtain corresponding picpointed coordinate (x _i, y _i);

Iii) calculate corresponding scanning coordinate by picpointed coordinate by the interior orientation transformation parameter

$\left[\begin{array}{c}I\\ J\end{array}\right]=\left[\begin{array}{cc}{m}_1& m_2\\ n_1& n_2\end{array}\right]\left[\begin{array}{c}x-x_0\\ y-y_0\end{array}\right]-\left[\begin{array}{c}{I}_0\\ J_0\end{array}\right]$

M in the formula ₁, m ₂, n ₁, n ₂, I ₀, J ₀Be the interior orientation parameter;

Iv) to the scanning coordinate of the pixel in each piece according to bilinearity polynomial interpolation correspondence

I＝a ₀+a ₁X+a ₂Y+a ₃XY

J＝b ₀+b ₁X+b ₂Y+b ₃XY

I in the formula, J are scanning coordinate, X, and Y is ground coordinate, a _i, b _iBe the bilinear transformation coefficient

V) gray scale interpolation

Four angle points of each bin are obtained its photo coordinate according to collinearity equation, be translated into scanning coordinate by the interior orientation parameter then, calculate its scanning coordinate for each piece inside according to the bilinear transformation pointwise, adopt the gray-scale value of each pixel of gray scale interpolating method interpolation again;

Vi) gray scale assignment

Each gray-scale value after the interpolation is composed one by one to each pixel after correcting;

Each point is promptly formed digital orthoimage after all handling;

(3) utilize digital terrain model, with even numbers sheet 2,4,6 ... make the supplementary plate of orthography, the process of making and the process of orthography are similar, step I just) projection function of introducing parallax method with the varied angle projection carries out projection and orthogonal projection when replacing making orthography, and the projection function of described varied angle projection introducing parallax method is $P_i=\frac{B{Z}_i}{H-Z_i},$ P in the formula _iBe the parallax of each point, Z _iBe the height value of each point, B is the photographic base of stereogram, the flying height when H is photography; Concrete making step is as follows:

By the square grid on the XY plane, get projecting direction according to the projection function of above-mentioned varied angle projection introducing parallax method and be parallel to XZ plane introducing parallax, according to the computing formula of the intersection point on varied angle projection ray and digital terrain model surface, with four angle points of graticule mesh project to digital elevation model obtain projection line and digital terrain model intersection point (X ' _iY ' _iZ ' _i), following calculation procedure goes on foot to vi with the ii that makes orthography;

(4) digital orthoimage that the odd number sheet is made is set into a seamless zone according to ground coordinate with it, writes down the regional extent of the digital orthoimage of each digital orthography after inlaying simultaneously;

(5) the scalable seamless stereoscopic model of point-to-point structure.

Images