CN105260610A - Multi-detector coordinate system transformation and error correction methods - Google Patents

Multi-detector coordinate system transformation and error correction methods Download PDF

Info

Publication number
CN105260610A
CN105260610A CN201510698671.4A CN201510698671A CN105260610A CN 105260610 A CN105260610 A CN 105260610A CN 201510698671 A CN201510698671 A CN 201510698671A CN 105260610 A CN105260610 A CN 105260610A
Authority
CN
China
Prior art keywords
amp
coordinate
step
coordinate system
prime
Prior art date
Application number
CN201510698671.4A
Other languages
Chinese (zh)
Other versions
CN105260610B (en
Inventor
金尔文
隋运峰
钟琦
李华琼
王雨果
鄢丹青
张中仅
Original Assignee
中国民用航空总局第二研究所
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 中国民用航空总局第二研究所 filed Critical 中国民用航空总局第二研究所
Priority to CN201510698671.4A priority Critical patent/CN105260610B/en
Publication of CN105260610A publication Critical patent/CN105260610A/en
Application granted granted Critical
Publication of CN105260610B publication Critical patent/CN105260610B/en

Links

Abstract

The invention relates to the field of multi-detector coordinate system transformation, in particular to multi-detector coordinate system transformation and error correction methods. Aiming at problems existing in the prior art, the invention provides the multi-detector coordinate system transformation and error correction methods. Aiming at the fact that an FOD detection system mainly depends on a radar detection technology, an image detection technology and a laser detection technology, the invention provides a mutual transformation mathematical model of detector coordinates and a runway global coordinate system and provides a method for calculating the parameters of the transformation model. On such basis, aiming at errors of transformation results, the invention provides the error correction method, so that the coordinate transformation accuracy is improved further. According to the invention, corresponding position information under the detector coordinate system of test points is measured through detectors; the parameters in the coordinate transformation model are solved according to the corresponding relationship; after target coordinates are obtained in the detection, a simplified formula for mutual transformation of detection coordinates of multiple detectors and the standard runway coordinates is calculated out.

Description

A kind of multidetector coordinate system transforms and error correction method

Technical field

The present invention relates to multidetector ordinate transform field, especially a kind of multidetector coordinate system transforms and error correction method.

Background technology

The nucleus equipment of fixed FOD detection system is the some probe units installed in runway both sides.Probe unit can be one or more equipment in radar, image and laser.After probe unit finds FOD target, the coordinates of targets under local Coordinate System can only be provided when lacking global information.But this coordinate is difficult to by intuitivism apprehension, be also difficult to by other equipment use, needs are transformed into can by the runway coordinate system of intuitivism apprehension.A coordinate system method for transformation is accurately that FOD detects a step crucial in holistic approach, and being the standard mutually changed of multiple probe unit coordinate and bridge, is also the basis of multidetector data fusion detection.

Some systems directly measure the position of probe unit, carry out displacement simply obtain runway coordinate to detector self coordinate.Such method is easy to operation, but result often exists larger conversion error, even exceedes the detecting error of detector.The way of science is, first sets up the mathematical model of ordinate transform, then solves the parameter in mathematical model, finally changes coordinate according to the parameter calculating gained.Simple transformation model is easy to calculating parameter, but transformation result error is larger; Otherwise complicated transformation model transformation result error is less, but is difficult to calculating parameter.Outstanding conversion method can calculate the parameter of transformation model by approach eaily, and transformation result accuracy is high.

Summary of the invention

Technical matters to be solved by this invention is: for above-mentioned Problems existing, a kind of multidetector coordinate system is provided to transform and error correction method, the present invention is directed to main radar, image and laser three kinds of Detection Techniques relied in FOD detection system and propose the mutual conversion mathematical model of detector self coordinate and runway global coordinate system, and propose the method calculating conversion model parameters and calculate.On this basis, for the error of transformation result, propose a kind of error correction method, further increase the precision of coordinate conversion.

The technical solution used in the present invention is as follows:

A kind of multidetector coordinate system method for transformation comprises:

Step 1: the transfer principle model proposing a kind of general detection coordinate system and runway coordinate system, the detection of a target be (θ at detector polar coordinates expression-form, d, δ), wherein d represents that detector is to being detected object distance, θ represents the position angle be detected between object and detector, and δ represents the angle of pitch be detected between object and detector; Under detector cartesian coordinate system, expression-form is P t(x t, y t, z t):

x T=dcosδcosθ

y T=dcosδsinθ

z T=dsinδ

By P t(x t, y t, z t) be transformed into the coordinate P of runway coordinate system g(x g, y g, z g) Coordinate Transformation Models:

P G=R 0P T+P 0

Wherein, R 0for rotation matrix, P 0(x 0, y 0, z 0) be the position of detector coordinates system initial point in runway coordinate system; Unknown parameter in this model is R 0, P 0, and detector is the whole of three components or wherein two according to the coordinate information that the difference of type obtains.

Step 2: on runway road surface, evenly N is set 1individually be detected object, be N 1individual test point, uses measuring instrument to record its coordinate under runway coordinate system and is respectively (x j, y j, z j); Use detector to record correspondence position information under the detector coordinates system of test point simultaneously; According to the parameter in corresponding relation solution procedure 1 Coordinate Transformation Models;

Step 3: obtain coordinates of targets in detection after, uses the parameter in step 2 in Formula of Coordinate System Transformation model to carry out abbreviation to step 1 Coordinate Transformation Models, and the detection coordinate and the runway standard coordinate that calculate multiple detector carry out the mutual formula of reduction changed.

Further, when described detector is radar, test point coordinate in radar detection gained radar fix system is distributed as (θ j, d j), 1≤j≤N 1, in step 1, Coordinate Transformation Models unknown parameter is δ, R 0, P 0; Namely as δ in actual measurement j, R 0, P 0for unknown number; Wherein δ jfor the angle of pitch be detected between object and radar that N1 test point is corresponding; θ jposition angle between the testee corresponding for N1 test point and radar; d jdistance between the testee corresponding for N1 test point and radar; The parameter solved in step 2 in Coordinate Transformation Models specifically comprises:

Step 21: structure A = 2 ( x 2 - x 1 ) 2 ( y 2 - y 1 ) 2 ( z 2 - z 1 ) 2 ( x 3 - x 2 ) 2 ( y 3 - y 2 ) 2 ( z 3 - z 2 ) ... ... ... 2 ( x N 1 - x N 1 - 1 ) 2 ( y N 1 - y N 1 - 1 ) 2 ( z N 1 - z N 1 - 1 ) ;

B = d 1 2 + x 2 2 + y 2 2 + z 2 2 - d 2 2 - x 1 2 - y 1 2 - z 1 2 d 2 2 + x 3 2 + y 3 2 + z 3 2 - d 3 2 - x 2 2 - y 2 2 - z 2 2 ... d N 1 - 1 2 + x N 1 2 + y N 1 2 + z N 1 2 - d N 1 2 - x N 1 - 1 2 - y N 1 - 1 2 - z N 1 - 1 2

According to calculating P 0=(A ta) -1a tb, obtains P 0;

Step 22: because R 0 -1(P g-P 0)=P t, note (x j-x 0, y j-y 0, z j-z 0) be (vx j, vy j, vz j), note R 0 -1for R 0 - 1 = r 11 r 12 r 13 r 12 r 22 r 23 r 13 r 23 r 33 ;

Structure t e m p = vx 1 vy 1 - vx 1 tanθ 1 vz 1 - vy 1 tanθ 1 - vz 1 tanθ 1 vx 2 vy 2 - vx 2 tanθ 2 vz 2 - vy 2 tanθ 2 - vz 2 tanθ 2 ... ... ... ... ... vx N 1 vy N 1 - vx N 1 tanθ N 1 vz N 1 - vy N 1 tanθ N 1 - vz N 1 tanθ N 1

And to temp ttemp carries out Eigenvalues Decomposition or svd, gets minimum non-zero eigenwert characteristic of correspondence vector, and this vector is designated as (r 11', r 12', r 13', r 22', r 23'), perform step 23;

Step 23: calculate:

r 11 = 2 r 11 ′ r 11 ′ 2 + 2 r 12 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2

r 12 = 2 r 12 ′ r 11 ′ 2 + 2 r 12 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2

r 13 = 2 r 13 ′ r 11 ′ 2 + 2 r 12 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2

r 22 = 2 r 22 ′ r 11 ′ 2 + 2 r 12 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2

r 23 = 2 r 23 ′ r 11 ′ 2 + 2 r 12 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2

r 33 = 1 - r 13 2 - r 23 2

Obtain R 0 -1in the value of all parameters, perform step 24:

Step 24: to R 0 -1finding the inverse matrix, obtains R 0value;

Step 25: all have z for target in any runway plane g≈ 0, and

R 0 - 1 ( x G y G 0 - P 0 ) = P T

I.e. r 11(x g-x 0)+r 12(y g-y 0)+r 13(-z 0)=x t, r 12(x g-x 0)+r 22(y g-y 0)+r 23(-z 0)=y t;

And have y T x T = r 12 ( x G - x 0 ) + r 22 ( y G - y 0 ) + r 23 ( - z 0 ) r 11 ( x G - x 0 ) + r 12 ( y G - y 0 ) + r 13 ( - z 0 ) = t a n θ ; x G 2+y G 2=d 2-z 0 2

According to above two restrictive conditions, order: k 1 = r 12 t a n θ - r 22 r 12 - r 11 t a n θ

k 2 = r 12 x 0 + r 22 y 0 + r 23 z 0 - tan θ ( r 11 x 0 + r 12 y 0 + r 13 z 0 ) r 12 - r 11 tan θ

Calculate y gfor linear equation in two unknowns (1+k 1 2) y g 2+ 2k 1k 2y g+ k 2 2-d 2+ z 0 2the normal solution of=0, and x g=k 1y g+ k 2

Thus when without the need to knowing parameter δ, realize the conversion of radar fix to runway coordinate.

Further, when described detector is radar, step 3 is respectively by step 31 or step 32 specific implementation:

Step 31: to coordinate P under the runway coordinate system that sets the goal g(x g, y g, z g), its radar detection coordinate (θ, d) computing method are:

P T = x T y T z T = R 0 - 1 ( P G - P 0 )

θ=artan(y T/x T)

d = x T 2 + y T 2 + z T 2 ;

Step 32: given target radar detection coordinate (θ, d), coordinate P under its runway coordinate system g(x g, y g, z g) computing method be: order

k 2 = r 12 x 0 + r 22 y 0 + r 23 z 0 - t a n θ ( r 11 x 0 + r 12 y 0 + r 13 z 0 ) r 12 - r 11 t a n θ ;

Obtain y gfor linear equation in two unknowns: (1+k 1 2) y g 2+ 2k 1k 2y g+ k 2 2-d 2+ z 0 2the normal solution of=0, and x g=k 1y g+ k 2, z g=0.

Further, when described detector is image detector, in step 1, Coordinate Transformation Models unknown parameter is d, R 0, P 0; Under image coordinate system, test point coordinate in camera detection gained image coordinate system is distributed as (θ j, δ j), 1≤j≤N 1; I.e. d in actual measurement j, R 0, P 0for unknown number; Wherein δ jfor the angle of pitch be detected between object and image detector that N1 test point is corresponding; θ jposition angle between the testee corresponding for N1 test point and image detector; d jdistance between the testee corresponding for N1 test point and image detector; The parameter solved in step 2 in Coordinate Transformation Models specifically comprises:

In image detection gained image coordinate system, coordinate is distributed as (θ j, δ j), 1≤j≤N 1, measuring detector coordinate under runway coordinate system is at P 0' centered by σ be radius, in the square space of 2 σ × 2, σ × 2 σ, get and be uniformly distributed the some P that spacing is τ 0'=(x 0', y 0', z 0') as P 0estimated value, x ~ 0 - σ ≤ x 0 ′ ≤ x ~ 0 + σ , y ~ 0 - σ ≤ y 0 ′ ≤ y ~ 0 + σ , z ~ 0 - σ ≤ z 0 ′ ≤ z ~ 0 + σ , The value of σ is 0.1 meter to 5 meters, and the value of τ is arrive to all P 0' perform following calculating:

Step 21: V is calculated to each test point j=(x j-x 0', y j-y 0', z j-z 0'), and by vectorial V jbe scaled vector of unit length (mould divided by self)

Step 22: W is calculated to each test point j=(cos δ jcos θ j, cos δ jsin θ j, sin δ j);

Step 23: compute vector K jfor the multiplication cross K of Wj and Vj j=W j× V j, be designated as K j=(kx j, ky j, kz j), and calculate ψ j=arsin (kz), ζ j=artan (ky j/ kx j);

Step 24: the angle between compute vector Wj and Vj

Step 25: ψ is calculated to all test points j, ζ j, the master sample difference of numeric distribution and ε, represent and calculate from x 1to x n1the master sample of N1 numeric distribution is poor altogether.

At all P 0' in, choose one group that ε value is minimum, order

P 0=P 0'

k x = c o s ( m e a n 1 ≤ j ≤ N 1 ( ψ j ) ) c o s ( m e a n 1 ≤ j ≤ N 1 ( ζ j ) )

k x = c o s ( m e a n 1 ≤ j ≤ N 1 ( ψ j ) ) sin ( m e a n 1 ≤ j ≤ N 1 ( ζ j ) )

k z = s i n ( m e a n 1 ≤ j ≤ N 1 ( ψ j ) )

R K = 0 - k z k y k z 0 - k x - k y k x 0

Obtain R 0, P 0, wherein represent and calculate from x 1to x n1the mean value of N1 numerical value altogether;

Step 26: order R 0 = t 11 t 12 t 13 t 12 t 22 t 23 t 13 t 23 t 33 , By P g=R 0p t+ P 0goal satisfaction relation t on known runway face 13dcos δ cos θ+t 23dcos δ sin θ+t 33dsin δ=-z 0;

Backwards calculation goes out parameter d: d = - z 0 t 13 c o s δ c o s θ + t 23 cos δ s i n θ + t 33 s i n δ .

Further, when described probe unit is image detector, step 3 specifically comprises:

Step 31: to coordinate P under the runway coordinate system that sets the goal g(x g, y g, z g), its image detection coordinate (θ, δ) computing method are

P T = x T y T z T = R 0 - 1 ( P G - P 0 )

δ = a r t a n ( z T / x T 2 + y T 2 )

θ=artan(y T/x T)

Given Graph as detection of a target coordinate (θ, δ), coordinate P under its runway coordinate system g(x g, y g, z g) computing method are

d = - z 0 t 13 c o s δ c o s θ + t 23 cos δ s i n θ + t 33 s i n δ

x T=dcosδcosθ

y T=dcosδsinθ

z T=dsinδ

According to P G = R 0 x T y T z T + P 0 , Obtain P g.

Further, when described probe unit is laser scanner, in step 1, Coordinate Transformation Models unknown parameter is R 0, P 0; Be detected object through laser acquisition, test point coordinate in laser coordinate system is distributed as (θ j, δ j, d j), 1≤j≤N 1; Wherein δ jfor the angle of pitch be detected between object and detector that N1 test point is corresponding; θ jposition angle between the testee corresponding for N1 test point and detector; d jdistance between the testee corresponding for N1 test point and detector; Then solve parameter in Coordinate Transformation Models in step 2 specifically to comprise:

Step 21: structure A = 2 ( x 2 - x 1 ) 2 ( y 2 - y 1 ) 2 ( z 2 - z 1 ) 2 ( x 3 - x 2 ) 2 ( y 3 - y 2 ) 2 ( z 3 - z 2 ) ... ... ... 2 ( x N 1 - x N 1 - 1 ) 2 ( y N 1 - y N 1 - 1 ) 2 ( z N 1 - z N 1 - 1 ) ;

B = d 1 2 + x 2 2 + y 2 2 + z 2 2 - d 2 2 - x 1 2 - y 1 2 - z 1 2 d 2 2 + x 3 2 + y 3 2 + z 3 2 - d 3 2 - x 2 2 - y 2 2 - z 2 2 ... d N 1 - 1 2 + x N 1 2 + y N 1 2 + z N 1 2 - d N 1 2 - x N 1 - 1 2 - y N 1 - 1 2 - z N 1 - 1 2

According to calculating P 0=(A ta) -1a tb, obtains P 0;

Step 22: V is calculated to each test point j=(x j-x 0', y j-y 0', z j-z 0'), and by vectorial V jbe scaled vector of unit length (mould divided by self)

Step 23: W is calculated to each test point j=(cos δ jcos θ j, cos δ jsin θ j, sin δ j);

Step 24: to all V jand W jcalculate average vector

V ‾ = 1 N 1 Σ j V j , W ‾ = 1 N 1 Σ j W j

Step 23: compute vector K is with multiplication cross and K is scaled vector of unit length

K = K | | K | |

Be designated as K=(kx, ky, kz), and compute vector with between angle

Step 24: structure

R K = 0 - k z k y k z 0 - k x - k y k x 0

Calculate

Obtain the R in model parameter 0.

Further, when described probe unit is laser, step 3 is specially:

Step 31: to coordinate P under the runway coordinate system that sets the goal g(x g, y g, z g), its laser acquisition coordinate (θ, δ, d) computing method are:

P T = x T y T z T = R 0 - 1 ( P G - P 0 )

δ = a r t a n ( z T / x T 2 + y T 2 )

θ=artan(y T/x T)

d = x T 2 + y T 2 + z T 2

Step 32: given laser acquisition coordinates of targets (θ, δ, d), coordinate P under its runway coordinate system g(x g, y g, z g) computing method are

x T=dcosδcosθ

y T=dcosδsinθ

z T=dsinδ

P G = R 0 x T y T z T + P 0 .

The error correction method that multidetector coordinate system transforms is P when calculating the coordinate of test point at runway coordinate system c(cx, cy, cz), and the coordinate of actual runway coordinate system is P g(gx, gy, gz), the error in X-axis Y direction is respectively Δ x=gx-cx, Δ y=gy-cy, time; Then the pass of error and coordinate is Δ x=f 1(cx, cy), Δ y=f 2(cx, cy), f 1, f 2, f 3, f 4be smooth surface function, comprise binary curved surface, dihydric phenol curved surface, binary cubic surface, trigonometric function, Gauss curved.

As described Δ x=f 1when (cx, cy) function is dihydric phenol curved surface, note N 1the coordinate that individual test point calculates runway coordinate system is respectively P c-j(cx j, cy j, cz j), 1≤j≤N 1, and actual coordinate is P g-j(gx j, gy j, gz j), 1≤j≤N 1, the error in this X-axis is Δ x j=gx j-cx j.Structural matrix:

A = cx 1 2 cx 1 cy 1 cy 1 2 cx 1 cy 1 1 cx 2 2 cx 2 cy 2 cy 2 2 cx 2 cy 2 1 ... ... ... ... ... ... cx N 1 2 cx N 1 cy N 1 cy N 1 2 cx N 1 cy N 1 1

B X = gx 1 - cx 1 gx 2 - cx 2 ... gx N 1 - cx N 1 B Y = gy 1 - cy 1 gy 2 - cy 2 ... gy N 1 - cy N 1

Make E x=(A ta) -1a tb x, E y=(A ta) -1a tb y;

And remember E=(e 1, e 2, e 3, e 4, e 5, e 6); H=(h 1, h 2, h 3, h 4, h 5, h 6)

Then the value after error correction calculating invocation point (cx, cy) be should be:

cx=cx+e 1cx 2+e 2cxcy+e 3cy 3+e 4cx+e 5cy+e 6

cy=cy+h 1cx 2+h 2cxcy+h 3cy 3+h 4cx+h 5cy+h 6

In sum, owing to have employed technique scheme, the invention has the beneficial effects as follows:

1, a kind of method that multidetector coordinate and runway coordinate are changed mutually is proposed, method employs complicated three-dimensional model, and the different parameters proposing this three-dimensional model under Given information different situations solves and simplifying method, compare common approximate model and can realize detector to less error and runway coordinate is changed mutually;

2, with runway coordinate for bridge, with the Formula of Coordinate System Transformation simplified for method, the conversion of coordinate between multiple detector can be realized fast;

3, can quantitative coordinates computed transformed error based on the error analysis of test point and correcting method, and reduce transformed error further on this basis.

Embodiment

All features disclosed in this instructions, or the step in disclosed all methods or process, except mutually exclusive feature and/or step, all can combine by any way.

Arbitrary feature disclosed in this instructions (comprising any accessory claim, summary), unless specifically stated otherwise, all can be replaced by other equivalences or the alternative features with similar object.That is, unless specifically stated otherwise, each feature is an example in a series of equivalence or similar characteristics.

Related description of the present invention:

1, runway coordinate system refers to the coordinate system defined with runway layout structure according to aerodrome standard, with the installation site of detector with towards irrelevant.

2, detector coordinates system refers to radar fix system, image coordinate system or laser coordinate system, and wherein radar fix system refers to radar is detector, and mark is detected the coordinate system of object location information; It is detector that image coordinate system refers to image detector, and mark is detected the coordinate system of object location information; It is detector that laser coordinate system refers to laser scanner, and mark is detected the coordinate system of object location information.Its coordinate origin and coordinate axis change because of installation site with towards difference, and survey instrument cannot be relied on to carry out high-precision measurement.

3, image detector refers to the camera be fixed on turntable.

Principle of work: the detection system that the present invention is directed to containing multiple detector exists the inconsistent problem of coordinate system, especially for radar, image and laser three kinds of Detection Techniques relied on main in FOD detection system, propose a kind of carrying out by respective coordinate system and runway coordinate system and mutually change mathematical model, and propose the method calculating conversion model parameters and calculate, and after drawing model parameter, simplify the Formula of Coordinate System Transformation of transformation model.On this basis, for the error of transformation result, propose a kind of error correction method, further increase the precision of coordinate conversion, also indirectly improve the precision of result of detection.

One, when detector is radar:

1, radar fix system transformation model

Under radar fix system, the coordinate expression-form of the detection of a target is (θ, d).Wherein d represents distance, and θ represents position angle.

Transformation model establishes radar fix system mid point P t(x t, y t, z t) be transformed into the coordinate P of runway coordinate system g(x g, y g, z g) formula is P g=R 0p t+ P 0

Wherein, R 0for rotation matrix, P 0(x 0, y 0, z 0) be the position of radar fix system initial point in runway coordinate system.And P t(x t, y t, z t) be pass through

x T=dcosδcosθ

y T=dcosδsinθ

z T=dsinδ

Calculate, δ is target pitch angle under radar fix system.Unknown parameter in this model is δ, R 0, P 0.

2, radar fix system conversion model parameters method for solving

Be uniformly distributed in detector monitoring range and put N 1individual test point (N 1value is generally between 9 to 85), its coordinate under runway coordinate system is respectively (x j, y j, z j), 1≤j≤N 1, in radar detection gained radar fix system, coordinate is distributed as (θ j, d j), 1≤j≤N 1, the unknown object angle of pitch is δ j.X jfor the coordinate of test point x-axis under runway coordinate system; y jfor the coordinate of test point y-axis under runway coordinate system; z jfor the coordinate of test point z-axis under runway coordinate system; X, y-axis are parallel with runway, and z-axis is vertical with runway, and x-axis and y-axis are mutually vertical;

By following calculation procedure computation model parameter:

Step 1: structure

A = 2 ( x 2 - x 1 ) 2 ( y 2 - y 1 ) 2 ( z 2 - z 1 ) 2 ( x 3 - x 2 ) 2 ( y 3 - y 2 ) 2 ( z 3 - z 2 ) ... ... ... 2 ( x N 1 - x N 1 - 1 ) 2 ( y N 1 - y N 1 - 1 ) 2 ( z N 1 - z N 1 - 1 )

B = d 1 2 + x 2 2 + y 2 2 + z 2 2 - d 2 2 - x 1 2 - y 1 2 - z 1 2 d 2 2 + x 3 2 + y 3 2 + z 3 2 - d 3 2 - x 2 2 - y 2 2 - z 2 2 ... d N 1 - 1 2 + x N 1 2 + y N 1 2 + z N 1 2 - d N 1 2 - x N 1 - 1 2 - y N 1 - 1 2 - z N 1 - 1 2

Calculate P 0=(A ta) -1a tb, obtains the P in model parameter 0;

Step 2: because R 0 -1(P g-P 0)=P t, note (x j-x 0, y j-y 0, z j-z 0) be (vx j, vy j, vz j), note R 0 -1for R 0 - 1 = r 11 r 12 r 13 r 12 r 22 r 23 r 13 r 23 r 33

Structure t e m p = vx 1 vy 1 - vx 1 tanθ 1 vz 1 - vy 1 tanθ 1 - vz 1 tanθ 1 vx 2 vy 2 - vx 2 tanθ 2 vz 2 - vy 2 tanθ 2 - vz 2 tanθ 2 ... ... ... ... ... vx N 1 vy N 1 - vx N 1 tanθ N 1 vz N 1 - vy N 1 tanθ N 1 - vz N 1 tanθ N 1

And to temp ttemp carries out Eigenvalues Decomposition or svd, gets minimum non-zero eigenwert characteristic of correspondence vector, and this vector is designated as (r 11', r 12', r 13', r 22', r 23');

Step 3: calculate

r 11 = 2 r 11 ′ r 11 ′ 2 + 2 r 12 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2

r 12 = 2 r 12 ′ r 11 ′ 2 + 2 r 12 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2

r 13 = 2 r 13 ′ r 11 ′ 2 + 2 r 12 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2

r 22 = 2 r 22 ′ r 11 ′ 2 + 2 r 12 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2

r 23 = 2 r 23 ′ r 11 ′ 2 + 2 r 12 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2

r 33 = 1 - r 13 2 - r 23 2

Obtain R 0 -1in the value of all parameters.

Step 4: to R 0 -1finding the inverse matrix, obtains R 0value.

Step 5: all have z for target in any runway plane g≈ 0, close like setting up below

R 0 - 1 ( x G y G 0 - P 0 ) = P T

I.e. r 11(x g-x 0)+r 12(y g-y 0)+r 13(-z 0)=x t, r 12(x g-x 0)+r 22(y g-y 0)+r 23(-z 0)=y t

And have y T x T = r 12 ( x G - x 0 ) + r 22 ( y G - y 0 ) + r 23 ( - z 0 ) r 11 ( x G - x 0 ) + r 12 ( y G - y 0 ) + r 13 ( - z 0 ) = t a n θ , x G 2+y G 2=d 2-z 0 2

According to above two restrictive conditions, order k 1 = r 12 t a n θ - r 22 r 12 - r 11 tan θ

k 2 = r 12 x 0 + r 22 y 0 + r 23 z 0 - t a n θ ( r 11 x 0 + r 12 y 0 + r 13 z 0 ) r 12 - r 11 t a n θ

Calculate y gfor linear equation in two unknowns (1+k 1 2) y g 2+ 2k 1k 2y g+ k 2 2-d 2+ z 0 2the normal solution of=0, and x g=k 1y g+ k 2

Thus when without the need to knowing parameter δ, realize the conversion of radar fix to runway coordinate.

3, radar fix system conversion formula

To coordinate P under the runway coordinate system that sets the goal g(x g, y g, z g), its radar detection coordinate (θ, d) computing method are

P T = x T y T z T = R 0 - 1 ( P G - P 0 )

θ=artan(y T/x T)

d = x T 2 + y T 2 + z T 2

Given target radar detection coordinate (θ, d), coordinate P under its runway coordinate system g(x g, y g, z g) computing method be: order k 1 = r 12 t a n θ - r 22 r 12 - r 11 tan θ

k 2 = r 12 x 0 + r 22 y 0 + r 23 z 0 - t a n θ ( r 11 x 0 + r 12 y 0 + r 13 z 0 ) r 12 - r 11 t a n θ

Y can be calculated gfor linear equation in two unknowns (1+k 1 2) y g 2+ 2k 1k 2y g+ k 2 2-d 2+ z 0 2the normal solution of=0, and x g=k 1y g+ k 2; z g=0

Two, when detector is image detector:

1, image coordinate system transformation model

Under image coordinate system, the coordinate expression-form of the detection of a target is (θ, δ).Wherein δ represents the angle of pitch, and θ represents position angle.

Transformation model establishes image coordinate system mid point P t(x t, y t, z t) be transformed into the coordinate P of runway coordinate system g(x g, y g, z g) formula is

P G=R 0P T+P 0

Wherein, R 0for rotation matrix, P 0(x 0, y 0, z 0) be the position of image coordinate system initial point in runway coordinate system.And P t(x t, y t, z t) be pass through

x T=dcosδcosθ

y T=dcosδsinθ

z T=dsinδ

Calculate, d is detector range-to-go.Unknown parameter in this model is d, R 0, P 0.

2, image coordinate system conversion model parameters method for solving

Be uniformly distributed in detector monitoring range and put N 1individual test point (N 1value is generally between 9 to 85), its coordinate under runway coordinate system is respectively (x j, y j, z j), 1≤j≤N 1, in image detection gained image coordinate system, coordinate is distributed as (θ j, δ j), 1≤j≤N 1.Rough measure equipment installation site coordinate under runway coordinate system is P ~ 0 = ( x ~ 0 , y ~ 0 , z ~ 0 ) .

At P 0' centered by σ be radius, in the square space of 2 σ × 2, σ × 2 σ, get and be uniformly distributed the some P that spacing is τ 0'=(x 0', y 0', z 0') as P 0estimated value, x ~ 0 - σ ≤ x 0 ′ ≤ x ~ 0 + σ , y ~ 0 - σ ≤ y 0 ′ ≤ y ~ 0 + σ , the value of σ is generally between 0.1 meter to 5 meters, and the value of τ is generally 1/100 to 1/10 of σ.To all P 0' perform following calculating:

Step 1: V is calculated to each test point j=(x j-x 0', y j-y 0', z j-z 0'), and by vectorial V jbe scaled vector of unit length (mould divided by self)

Step 2: W is calculated to each test point j=(cos δ jcos θ j, cos δ jsin θ j, sin δ j);

Step 3: compute vector K jfor the multiplication cross K of Wj and Vj j=W j× V j, be designated as K j=(kx j, ky j, kz j), and calculate ψ j=arsin (kz), ζ j=artan (ky j/ kx j);

Step 4: the angle between compute vector Wj and Vj

Step 5: ψ is calculated to all test points j, ζ jj, ζ jthe angle of pitch and horizontal angle for vectorial Kj), the master sample difference of numeric distribution and ε, represent and calculate from x 1to x n1the master sample of N1 numeric distribution is poor altogether.

At all P 0' in, choose one group that ε value is minimum, order

P 0=P 0'

k x = c o s ( m e a n 1 ≤ j ≤ N 1 ( ψ j ) ) c o s ( m e a n 1 ≤ j ≤ N 1 ( ζ j ) )

k x = c o s ( m e a n 1 ≤ j ≤ N 1 ( ψ j ) ) sin ( m e a n 1 ≤ j ≤ N 1 ( ζ j ) )

k z = s i n ( m e a n 1 ≤ j ≤ N 1 ( ψ j ) )

R K = 0 - k z k y k z 0 - k x - k y k x 0

Obtain the R in model parameter 0, P 0, wherein represent and calculate from x 1to x n1the mean value of N1 numerical value altogether.

Step 6: order R 0 = t 11 t 12 t 13 t 12 t 22 t 23 t 13 t 23 t 33 , By P g=R 0p t+ P 0relation below goal satisfaction on known runway face

t 13dcosδcosθ+t 23dcosδsinθ+t 33dsinδ=-z 0

Backwards calculation goes out parameter d

d = - z 0 t 13 c o s δ c o s θ + t 23 cos δ sin θ + t 33 s i n δ

3, image coordinate system conversion formula

To coordinate P under the runway coordinate system that sets the goal g(x g, y g, z g), its image detection coordinate (θ, δ) computing method are

P T = x T y T z T = R 0 - 1 ( P G - P 0 )

δ = a r t a n ( z T / x T 2 + y T 2 )

θ=artan(y T/x T)

Given Graph as detection of a target coordinate (θ, δ), coordinate P under its runway coordinate system g(x g, y g, z g) computing method are

d = - z 0 t 13 c o s δ c o s θ + t 23 cos δ s i n θ + t 33 s i n δ

x T=dcosδcosθ

y T=dcosδsinθ

z T=dsinδ

P G = R 0 x T y T z T + P 0

Three, when detector is laser scanner:

1, laser coordinate system transformation model

Under laser coordinate system, the coordinate expression-form of the detection of a target is (θ, δ, d).Wherein d represents distance, and θ represents position angle, and δ represents the angle of pitch.

Transformation model establishes laser coordinate system mid point P t(x t, y t, z t) be transformed into the coordinate P of runway coordinate system g(x g, y g, z g) formula is

P G=R 0P T+P 0

Wherein, R 0for rotation matrix, P 0(x 0, y 0, z 0) be the position of laser coordinate system initial point in runway coordinate system.And P t(x t, y t, z t) be pass through

x T=dcosδcosθ

y T=dcosδsinθ

z T=dsinδ

Calculate.Unknown parameter in this model is R 0, P 0.

2, laser coordinate system conversion model parameters method for solving

Be uniformly distributed in detector monitoring range and put N 1individual test point (N 1value is generally between 9 to 85), its coordinate under runway coordinate system is respectively (x j, y j, z j), 1≤j≤N 1, in laser acquisition gained laser coordinate system, coordinate is distributed as (θ j, δ j, d j), 1≤j≤N 1.

By following calculation procedure computation model parameter:

Step 1: structure

A = 2 ( x 2 - x 1 ) 2 ( y 2 - y 1 ) 2 ( z 2 - z 1 ) 2 ( x 3 - x 2 ) 2 ( y 3 - y 2 ) 2 ( z 3 - z 2 ) ... ... ... 2 ( x N 1 - x N 1 - 1 ) 2 ( y N 1 - y N 1 - 1 ) 2 ( z N 1 - z N 1 - 1 )

B = d 1 2 + x 2 2 + y 2 2 + z 2 2 - d 2 2 - x 1 2 - y 1 2 - z 1 2 d 2 2 + x 3 2 + y 3 2 + z 3 2 - d 3 2 - x 2 2 - y 2 2 - z 2 2 ... d N 1 - 1 2 + x N 1 2 + y N 1 2 + z N 1 2 - d N 1 2 - x N 1 - 1 2 - y N 1 - 1 2 - z N 1 - 1 2

Calculate P 0=(A ta) -1a tb, obtains the P in model parameter 0;

Step 2: V is calculated to each test point j=(x j-x 0', y j-y 0', z j-z 0'), and by vectorial V jbe scaled vector of unit length (mould divided by self)

Step 3: W is calculated to each test point j=(cos δ jcos θ j, cos δ jsin θ j, sin δ j);

Step 4: to all V jand W jcalculate average vector

V ‾ = 1 N 1 Σ j V j , W ‾ = 1 N 1 Σ j W j

Step 5: compute vector K is with multiplication cross and K is scaled vector of unit length

K = K | | K | |

Be designated as K=(kx, ky, kz), and compute vector with between angle

Step 6: structure R K = 0 - k z k y k z 0 - k x - k y k x 0

Calculate

Obtain the R in model parameter 0.

3, laser coordinate system conversion formula

To coordinate P under the runway coordinate system that sets the goal g(x g, y g, z g), its laser acquisition coordinate (θ, δ, d) computing method are

P T = x T y T z T = R 0 - 1 ( P G - P 0 )

δ = a r t a n ( z T / x T 2 + y T 2 )

θ=artan(y T/x T)

d = x T 2 + y T 2 + z T 2

Given laser acquisition coordinates of targets (θ, δ, d), coordinate P under its runway coordinate system g(x g, y g, z g) computing method are

x T=dcosδcosθ

y T=dcosδsinθ

z T=dsinδ

P G = R 0 x T y T z T + P 0

Four, error correction method

Due to many reasons, measuring error, the detector actual configuration of such as detector itself are more more complicated than the model proposed in the present invention, there is error runway pavement and ideal plane, and the coordinate using aforementioned conversion method to generate and actual coordinate still exist certain error.

It is P that design calculates coordinate c(cx, cy, cz), and actual coordinate is P g(gx, gy, gz).Error in X-axis Y direction is respectively Δ x=gx-cx, Δ y=gy-cy, think that the pass of error and coordinate is Δ x=f 1(cx, cy), Δ y=f 2(cx, cy), f 1, f 2, f 3, f 4be smooth surface function, including but not limited to binary curved surface, dihydric phenol curved surface, binary cubic surface, trigonometric function, Gauss curved.Error calculation method can be obtained by solving smooth surface parameter.

With Δ x=f 1(cx, cy) function is dihydric phenol curved surface is example, note N 1the coordinate that individual test point calculates is respectively P c-j(cx j, cy j, cz j), 1≤j≤N 1, and actual coordinate is P g-j(gx j, gy j, gz j), 1≤j≤N 1, the error in this X-axis is Δ x j=gx j-cx j.Structural matrix

A = cx 1 2 cx 1 cy 1 cy 1 2 cx 1 cy 1 1 cx 2 2 cx 2 cy 2 cy 2 2 cx 2 cy 2 1 ... ... ... ... ... ... cx N 1 2 cx N 1 cy N 1 cy N 1 2 cx N 1 cy N 1 1

B X = gx 1 - cx 1 gx 2 - cx 2 ... gx N 1 - cx N 1 B Y = gy 1 - cy 1 gy 2 - cy 2 ... gy N 1 - cy N 1

Make E x=(A ta) -1a tb x, E y=(A ta) -1a tb y

And remember E=(e 1, e 2, e 3, e 4, e 5, e 6), H=(h 1, h 2, h 3, h 4, h 5, h 6)

Then the value after error correction calculating invocation point (cx, cy) be should be:

cx=cx+e 1cx 2+e 2cxcy+e 3cy 3+e 4cx+e 5cy+e 6

cy=cy+h 1cx 2+h 2cxcy+h 3cy 3+h 4cx+h 5cy+h 6

The present invention is not limited to aforesaid embodiment.The present invention expands to any new feature of disclosing in this manual or any combination newly, and the step of the arbitrary new method disclosed or process or any combination newly.

Claims (9)

1. a multidetector coordinate system method for transformation, is characterized in that comprising:
Step 1: the transfer principle model proposing a kind of general detection coordinate system and runway coordinate system, the detection of a target be (θ at detector polar coordinates expression-form, d, δ), wherein d represents that detector is to being detected object distance, θ represents the position angle be detected between object and detector, and δ represents the angle of pitch be detected between object and detector; Under detector cartesian coordinate system, expression-form is P t(x t, y t, z t):
x T=dcosδcosθ
y T=dcosδsinθ
z T=dsinδ
By P t(x t, y t, z t) be transformed into the coordinate P of runway coordinate system g(x g, y g, z g) Coordinate Transformation Models:
P G=R 0P T+P 0
Wherein, R 0for rotation matrix, P 0(x 0, y 0, z 0) be the position of detector coordinates system initial point in runway coordinate system; Unknown parameter in this model is R 0, P 0, and detector is the whole of three components or wherein two according to the coordinate information that the difference of type obtains.
Step 2: on runway road surface, evenly N is set 1individually be detected object, be N 1individual test point, uses measuring instrument to record its test point coordinate under runway coordinate system and is respectively (x j, y j, z j); Use detector to record test point correspondence position information under detector coordinates system simultaneously; According to the parameter in corresponding relation solution procedure 1 Coordinate Transformation Models;
Step 3: obtain coordinates of targets in detection after, uses the parameter in step 2 in Formula of Coordinate System Transformation model to carry out abbreviation to step 1 Coordinate Transformation Models, and the detection coordinate and the runway standard coordinate that calculate multiple detector carry out the mutual formula changed.
2. a kind of multidetector coordinate system method for transformation according to claim 1, is characterized in that when described detector is radar, and test point coordinate in radar detection gained radar fix system is distributed as (θ j, d j), 1≤j≤N 1, in step 1, Coordinate Transformation Models unknown parameter is δ, R 0, P 0; Namely as δ in actual measurement j, R 0, P 0for unknown number; Wherein δ jfor the angle of pitch be detected between object and radar that N1 test point is corresponding; θ jposition angle between the testee corresponding for N1 test point and radar; d jthe testee corresponding for N1 test point and between distance; The parameter solved in step 2 in Coordinate Transformation Models specifically comprises:
Step 21: structure A = 2 ( x 2 - x 1 ) 2 ( y 2 - y 1 ) 2 ( z 2 - z 1 ) 2 ( x 3 - x 2 ) 2 ( y 3 - y 2 ) 2 ( z 3 - z 2 ) ... ... ... 2 ( x N 1 - x N 1 - 1 ) 2 ( y N 1 - y N 1 - 1 ) 2 ( z N 1 - z N 1 - 1 ) ;
B = d 1 2 + x 2 2 + y 2 2 + z 2 2 - d 2 2 - x 1 2 - y 1 2 - z 1 2 d 2 2 + x 3 2 + y 3 2 + z 3 2 - d 3 2 - x 2 2 - y 2 2 - z 2 2 ... d N 1 - 1 2 + x N 1 2 + y N 1 2 + z N 1 2 - d N 1 2 - x N 1 - 1 2 - y N 1 - 1 2 - z N 1 - 1 2
According to calculating P 0=(A ta) -1a tb, obtains P 0;
Step 22: because R 0 -1(P g-P 0)=P t, note (x j-x 0, y j-y 0, z j-z 0) be (vx j, vy j, vz j), note R 0 -1for
R 0 - 1 = r 11 r 12 r 13 r 12 r 22 r 23 r 13 r 23 r 33 ;
Structure t e m p = vx 1 vy 1 - vx 1 tanθ 1 vz 1 - vy 1 tanθ 1 - vz 1 tanθ 1 vx 2 vy 2 - vx 2 tanθ 2 vz 2 - vy 2 tanθ 2 - vz 2 tanθ 2 ... ... ... ... ... vx N 1 vy N 1 - vx N 1 tanθ N 1 vz N 1 - vy N 1 tanθ N 1 - vz N 1 tanθ N 1
And to temp ttemp carries out Eigenvalues Decomposition or svd, gets minimum non-zero eigenwert characteristic of correspondence vector, and this vector is designated as (r 11', r 12', r 13', r 22', r 23'), perform step 23;
Step 23: calculate:
r 11 = 2 r 11 ′ r 11 ′ 2 + 2 r 1 2 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2
r 12 = 2 r 12 ′ r 11 ′ 2 + 2 r 1 2 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2
r 13 = 2 r 13 ′ r 11 ′ 2 + 2 r 1 2 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2
r 22 = 2 r 22 ′ r 11 ′ 2 + 2 r 1 2 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2
r 23 = 2 r 23 ′ r 11 ′ 2 + 2 r 1 2 ′ 2 + r 13 ′ 2 + r 22 ′ 2 + r 23 ′ 2
r 33 = 1 - r 13 2 - r 23 2
Obtain R 0 -1in the value of all parameters, perform step 24:
Step 24: to R 0 -1finding the inverse matrix, obtains R 0value;
Step 25: all have z for target in any runway plane g≈ 0, and
R 0 - 1 ( x G y G 0 - P 0 ) = P T
I.e. r 11(x g-x 0)+r 12(y g-y 0)+r 13(-z 0)=x t;
r 12(x G-x 0)+r 22(y G-y 0)+r 23(-z 0)=y T
And have y T x T = r 12 ( x G - x 0 ) + r 22 ( y G - y 0 ) + r 23 ( - z 0 ) r 11 ( x G - x 0 ) + r 12 ( y G - y 0 ) + r 13 ( - z 0 ) = t a n θ
x G 2+y G 2=d 2-z 0 2
According to above two restrictive conditions, order:
k 2 = r 12 x 0 + r 22 y 0 + r 23 z 0 - t a n θ ( r 11 x 0 + r 12 y 0 + r 13 z 0 ) r 12 - r 11 t a n θ
Calculate y gfor linear equation in two unknowns (1+k 1 2) y g 2+ 2k 1k 2y g+ k 2 2-d 2+ z 0 2the normal solution of=0, and x g=k 1y g+ k 2
Thus when without the need to knowing parameter δ, realize the conversion of radar fix to runway coordinate.
3. a kind of multidetector coordinate system method for transformation according to claim 2, is characterized in that, when described detector is radar, step 3 is respectively by step 31 or step 32 specific implementation:
Step 31: to coordinate P under the runway coordinate system that sets the goal g(x g, y g, z g), its radar detection coordinate (θ, d) computing method are:
P T = x T y T z T = R 0 - 1 ( P G - P 0 )
θ=artan(y T/x T)
d = x T 2 + y T 2 + z T 2 ;
Step 32: given target radar detection coordinate (θ, d), coordinate P under its runway coordinate system g(x g, y g, z g) computing method be: order
k 2 = r 12 x 0 + r 22 y 0 + r 23 z 0 - t a n θ ( r 11 x 0 + r 12 y 0 + r 13 z 0 ) r 12 - r 11 t a n θ ;
Obtain y gfor linear equation in two unknowns: (1+k 1 2) y g 2+ 2k 1k 2y g+ k 2 2-d 2+ z 0 2the normal solution of=0, and x g=k 1y g+ k 2; z g=0.
4. a kind of multidetector coordinate system method for transformation according to claim 1, is characterized in that, when described detector is image detector, in step 1, Coordinate Transformation Models unknown parameter is d, R 0, P 0; Under image coordinate system, test point coordinate in camera detection gained image coordinate system is distributed as (θ j, δ j), 1≤j≤N 1; I.e. d in actual measurement j, R 0, P 0for unknown number; Wherein δ jfor the angle of pitch be detected between object and image detector that N1 test point is corresponding; θ jposition angle between the testee corresponding for N1 test point and image detector; d jdistance between the testee corresponding for N1 test point and image detector; The parameter solved in step 2 in Coordinate Transformation Models specifically comprises: in image detector gained image coordinate system, coordinate is distributed as (θ j, δ j), 1≤j≤N 1, measuring detector coordinate under runway coordinate system is at P 0' centered by σ be radius, in the square space of 2 σ × 2, σ × 2 σ, get and be uniformly distributed the some P that spacing is τ 0'=(x 0', y 0', z 0') as P 0estimated value, the value of σ is 0.1 meter to 5 meters, and the value of τ is arrive to all P 0' perform following calculating:
Step 21: V is calculated to each test point j=(x j-x 0', y j-y 0', z j-z 0'), and by vectorial V jbe scaled vector of unit length (mould divided by self)
Step 22: W is calculated to each test point j=(cos δ jcos θ j, cos δ jsin θ j, sin δ j);
Step 23: compute vector K jfor the multiplication cross K of Wj and Vj j=W j× V j, be designated as K j=(kx j, ky j, kz j), and calculate ψ j=arsin (kz), ζ j=artan (ky j/ kx j);
Step 24: the angle between compute vector Wj and Vj
Step 25: ψ is calculated to all test points j, ζ j, the master sample difference of numeric distribution and ε, represent and calculate from x 1to x n1the master sample of N1 numeric distribution is poor altogether; ψ j, ζ jfor the angle of pitch and the horizontal angle of vectorial Kj;
At all P 0' in, choose one group that ε value is minimum, order
P 0=P 0'
k x = c o s ( m e a n 1 ≤ j ≤ N 1 ( ψ j ) ) c o s ( m e a n 1 ≤ j ≤ N 1 ( ζ j ) )
k x = c o s ( m e a n 1 ≤ j ≤ N 1 ( ψ j ) ) sin ( m e a n 1 ≤ j ≤ N 1 ( ζ j ) )
k z = s i n ( m e a n 1 ≤ j ≤ N 1 ( ψ j ) )
R K = 0 - k z k y k z 0 - k x - k y k x 0
Obtain R 0, P 0, wherein represent and calculate from x 1to x n1the mean value of N1 numerical value altogether;
Step 26: order R 0 = t 11 t 12 t 13 t 12 t 22 t 23 t 13 t 23 t 33 , By P g=R 0p t+ P 0goal satisfaction relation: t on known runway face 13dcos δ cos θ+t 23dcos δ sin θ+t 33dsin δ=-z 0;
Backwards calculation goes out parameter d: d = - z 0 t 13 c o s δ c o s θ + t 23 cos δ s i n θ + t 33 s i n δ .
5. a kind of multidetector coordinate system method for transformation according to claim 1, is characterized in that step 3 specifically comprises when described probe unit is image detector:
Step 31: to coordinate P under the runway coordinate system that sets the goal g(x g, y g, z g), its image detection coordinate (θ, δ) computing method are
P T = x T y T z T = R 0 - 1 ( P G - P 0 ) ;
δ = a r t a n ( z T / x T 2 + y T 2 ) ;
θ=artan(y T/x T);
Given Graph as detection of a target coordinate (θ, δ), coordinate P under its runway coordinate system g(x g, y g, z g) computing method are
d = - z 0 t 13 c o s δ c o s θ + t 23 cos δ s i n θ + t 33 s i n δ
x T=dcosδcosθ
y T=dcosδsinθ
z T=dsinδ
According to P G = R 0 x T y T z T + P 0 , Obtain P g.
6. a kind of multidetector coordinate system method for transformation according to claim 1, is characterized in that, when described detector is laser scanner, in step 1, Coordinate Transformation Models unknown parameter is R 0, P 0; Be detected object through laser acquisition, test point coordinate in laser coordinate system is distributed as (θ j, δ j, d j), 1≤j≤N 1; Wherein δ jfor the angle of pitch be detected between object and laser scanner that N1 test point is corresponding; θ jposition angle between the testee corresponding for N1 test point and laser scanner; d jdistance between the testee corresponding for N1 test point and laser scanner; Then solve parameter in Coordinate Transformation Models in step 2 specifically to comprise:
Step 21: structure A = 2 ( x 2 - x 1 ) 2 ( y 2 - y 1 ) 2 ( z 2 - z 1 ) 2 ( x 3 - x 2 ) 2 ( y 3 - y 2 ) 2 ( z 3 - z 2 ) ... ... ... 2 ( x N 1 - x N 1 - 1 ) 2 ( y N 1 - y N 1 - 1 ) 2 ( z N 1 - z N 1 - 1 ) ;
B = d 1 2 + x 2 2 + y 2 2 + z 2 2 - d 2 2 - x 1 2 - y 1 2 - z 1 2 d 2 2 + x 3 2 + y 3 2 + z 3 2 - d 3 2 - x 2 2 - y 2 2 - z 2 2 ... d N 1 - 1 2 + x N 1 2 + y N 1 2 + z N 1 2 - d N 1 2 - x N 1 - 1 2 - y N 1 - 1 2 - z N 1 - 1 2
According to calculating P 0=(A ta) -1a tb, obtains P 0;
Step 22: V is calculated to each test point j=(x j-x 0', y j-y 0', z j-z 0'), and by vectorial V jbe scaled vector of unit length (mould divided by self)
Step 23: W is calculated to each test point j=(cos δ jcos θ j, cos δ jsin θ j, sin δ j);
Step 24: to all V jand W jcalculate average vector
V ‾ = 1 N 1 Σ j V j , W ‾ = 1 N 1 Σ j W j
Step 23: compute vector K is with multiplication cross and K is scaled vector of unit length
K = K | | K | |
Be designated as K=(kx, ky, kz), and compute vector with between angle
Step 24: structure R K = 0 - k z k y k z 0 - k x - k y k x 0
Calculate
Obtain the R in model parameter 0.
7. a kind of multidetector coordinate system method for transformation according to claim 1, is characterized in that step 3 is specially when described probe unit is laser scanner:
Step 31: to coordinate P under the runway coordinate system that sets the goal g(x g, y g, z g), its laser acquisition coordinate (θ, δ, d) computing method are:
P T = x T y T z T = R 0 - 1 ( P G - P 0 )
δ = a r t a n ( z T / x T 2 + y T 2 )
θ=artan(y T/x T)
d = x T 2 + y T 2 + z T 2
Step 32: given laser acquisition coordinates of targets (θ, δ, d), coordinate P under its runway coordinate system g(x g, y g, z g) computing method are
x T=dcosδcosθ
y T=dcosδsinθ
z T=dsinδ
P G = R 0 x T y T z T + P 0 .
8. an error correction method for multidetector coordinate system conversion, is characterized in that when calculating test point at the coordinate of runway coordinate system be P c(cx, cy, cz), and the coordinate of actual runway coordinate system is P g(gx, gy, gz), the error in X-axis Y direction is respectively Δ x=gx-cx, Δ y=gy-cy, time; Then the pass of error and coordinate is Δ x=f 1(cx, cy), Δ y=f 2(cx, cy), f 1, f 2, f 3, f 4be smooth surface function, comprise binary curved surface, dihydric phenol curved surface, binary cubic surface, trigonometric function, Gauss curved.
9. the error correction method of a kind of multidetector coordinate system conversion according to claim 8, is characterized in that Δ x=f 1when (cx, cy) function is dihydric phenol curved surface, note N 1the coordinate that individual test point calculates runway coordinate system is respectively P c-j(cx j, cy j, cz j), 1≤j≤N 1, and actual coordinate is P g-j(gx j, gy j, gz j), 1≤j≤N 1, the error in this X-axis is Δ x j=gx j-cx j; Structural matrix:
A = cx 1 2 cx 1 cy 1 cy 1 2 cx 1 cy 1 1 cx 2 2 cx 2 cy 2 cy 2 2 cx 2 cy 2 1 ... ... ... ... ... ... cx N 1 2 cx N 1 cy N 1 cy N 1 2 cx N 1 cy N 1 1
B X = g x 1 - c x 1 gx 2 - cx 2 ... g x N 1 - c x N 1 B Y = g y 1 - c y 1 gy 2 - cy 2 ... g y N 1 - c y N 1
Make E x=(A ta) -1a tb x; E y=(A ta) -1a tb y
And remember E=(e 1, e 2, e 3, e 4, e 5, e 6), H=(h 1, h 2, h 3, h 4, h 5, h 6);
Then the value after error correction calculating invocation point (cx, cy) be should be:
cx=cx+e 1cx 2+e 2cxcy+e 3cy 3+e 4cx+e 5cy+e 6
cy=cy+h 1cx 2+h 2cxcy+h 3cy 3+h 4cx+h 5cy+h 6
CN201510698671.4A 2015-10-26 2015-10-26 A kind of multi-detector coordinate system conversion and error correction method CN105260610B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510698671.4A CN105260610B (en) 2015-10-26 2015-10-26 A kind of multi-detector coordinate system conversion and error correction method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510698671.4A CN105260610B (en) 2015-10-26 2015-10-26 A kind of multi-detector coordinate system conversion and error correction method

Publications (2)

Publication Number Publication Date
CN105260610A true CN105260610A (en) 2016-01-20
CN105260610B CN105260610B (en) 2018-01-23

Family

ID=55100299

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510698671.4A CN105260610B (en) 2015-10-26 2015-10-26 A kind of multi-detector coordinate system conversion and error correction method

Country Status (1)

Country Link
CN (1) CN105260610B (en)

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH06249954A (en) * 1993-02-26 1994-09-09 Toshiba Corp Radar signal processor
US5937102A (en) * 1996-10-09 1999-08-10 California Institute Of Technology Image reconstruction
US6064388A (en) * 1997-11-10 2000-05-16 Cognex Corporation Cartesian to polar coordinate transformation
WO2007073751A1 (en) * 2005-12-23 2007-07-05 Barco Orthogon Gmbh Radar scan converter and method for transforming
CN100367046C (en) * 1999-11-29 2008-02-06 密执安特种矿石公司 Measurement of wear of fireproof lining of metallurgical vessel
WO2008038889A1 (en) * 2006-09-29 2008-04-03 Electronics And Telecommunications Research Institute Radar tracking device and method thereof
CN100498370C (en) * 2007-06-25 2009-06-10 北京航空航天大学 Error calibrating method for high dynamic, multivariate and asynchronous nonitoring system
CN103577697A (en) * 2013-11-12 2014-02-12 中国民用航空总局第二研究所 FOD detection method based on road surface point cloud data
CN104658026A (en) * 2015-02-15 2015-05-27 北京科艺有容科技有限责任公司 Blend shape method based on polar coordinate system

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH06249954A (en) * 1993-02-26 1994-09-09 Toshiba Corp Radar signal processor
US5937102A (en) * 1996-10-09 1999-08-10 California Institute Of Technology Image reconstruction
US6064388A (en) * 1997-11-10 2000-05-16 Cognex Corporation Cartesian to polar coordinate transformation
CN100367046C (en) * 1999-11-29 2008-02-06 密执安特种矿石公司 Measurement of wear of fireproof lining of metallurgical vessel
WO2007073751A1 (en) * 2005-12-23 2007-07-05 Barco Orthogon Gmbh Radar scan converter and method for transforming
WO2008038889A1 (en) * 2006-09-29 2008-04-03 Electronics And Telecommunications Research Institute Radar tracking device and method thereof
CN100498370C (en) * 2007-06-25 2009-06-10 北京航空航天大学 Error calibrating method for high dynamic, multivariate and asynchronous nonitoring system
CN103577697A (en) * 2013-11-12 2014-02-12 中国民用航空总局第二研究所 FOD detection method based on road surface point cloud data
CN104658026A (en) * 2015-02-15 2015-05-27 北京科艺有容科技有限责任公司 Blend shape method based on polar coordinate system

Non-Patent Citations (6)

* Cited by examiner, † Cited by third party
Title
卢发兴,等: "直角坐标系统中的误差误差相关函数解析分析", 《海军工程大学学报》 *
吴振亚,等;: "一种基于ADS-B的雷达误差实时融合校正算法", 《西南交通大学学报》 *
周智: "基于激光雷达的智能车主动安全算法与模型研究", 《中国优秀硕士学位论文全文数据库信息科技辑(月刊)》 *
张明旭,主编;: "《矿井地质理论与技术研究新进展2012年11月第1版》", 30 November 2012 *
王新刚,: "遥感影像几何校正精度分析", 《中国优秀硕士学位论文全文数据库基础科学辑(月刊)》 *
郑莎莎: "基于激光雷达点云数据的树冠结构三维空间异质性分析", 《中国优秀硕士学位论文全文数据库信息科技辑(月刊)》 *

Also Published As

Publication number Publication date
CN105260610B (en) 2018-01-23

Similar Documents

Publication Publication Date Title
Kurazume et al. Cooperative positioning with multiple robots
US7932718B1 (en) System and method using magnetic anomaly field magnitudes for detection, localization, classification and tracking of magnetic objects
Müller On the accuracy of turbulence measurements with inclined hot wires
Cooray et al. Large-scale bulk motions complicate the Hubble diagram
CN101825717B (en) Carrier smoothing code pseudorange technology-based dynamic attitude positioning method
Han et al. Monitoring tunnel profile by means of multi-epoch dispersed 3-D LiDAR point clouds
CH697492B1 (en) A method for estimating the accuracy of azimuthal orientations, methods for providing azimuthal orientation of a portable target device and an estimated value of their accuracy and portable target device.
CN104374385B (en) A kind of new method of seabed array of magnetic sensors target positioning
Raquet et al. Use of a Covariance Analysis Technique for Predicting Performance of Regional‐Area Differential Code and Carrier‐Phase Networks
CN104021676B (en) Vehicle location based on vehicle dynamic video features and vehicle speed measurement method
RU2432580C1 (en) Method to determine coordinates of radio-wave radiation source in process of amplitude-phase direction finding on board of aircraft
Muelaner et al. Estimation of uncertainty in three-dimensional coordinate measurement by comparison with calibrated points
CN103323855B (en) A kind of precision acquisition methods of baseline dynamic measurement system
CN104316903B (en) A kind of three station positioning using TDOA performance test appraisal procedures
White The comparison plot
Lynch et al. Material acceleration estimation by four-pulse tomo-PIV
CN102788578B (en) Matching navigation method based on local gravity field approximation
Bian et al. Study on GPS attitude determination system aided INS using adaptive Kalman filter
CN103744052B (en) A kind of double star being applied to aerial target location surveys time-of-arrival direction finding method and device
CN106289246B (en) A kind of flexible link arm measure method based on position and orientation measurement system
CN103616035B (en) A kind of performance parameter calibration method of laser strapdown inertial navigation system
CN104729506A (en) Unmanned aerial vehicle autonomous navigation positioning method with assistance of visual information
CA2866174C (en) Fault detection for pipelines
CN102073035B (en) Method for evaluating simulation performance of spaceborne interferometric synthetic aperture radar system
CN104729537B (en) A kind of in-orbit real-time compensation method of star sensor low frequency aberration

Legal Events

Date Code Title Description
PB01 Publication
C06 Publication
SE01 Entry into force of request for substantive examination
C10 Entry into substantive examination
GR01 Patent grant
GR01 Patent grant