CN105631063B - A kind of moment forecasting procedure that crosses based on target signature - Google Patents
A kind of moment forecasting procedure that crosses based on target signature Download PDFInfo
- Publication number
- CN105631063B CN105631063B CN201410596963.2A CN201410596963A CN105631063B CN 105631063 B CN105631063 B CN 105631063B CN 201410596963 A CN201410596963 A CN 201410596963A CN 105631063 B CN105631063 B CN 105631063B
- Authority
- CN
- China
- Prior art keywords
- platform
- observed
- crosses
- moment
- observation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Image Analysis (AREA)
Abstract
The invention belongs to utilize radio wave ranging technology field, and in particular to a kind of moment forecasting procedure that crosses based on target signature；The geometries characteristic information, gray feature information and geometric area characteristic information for being observed platform are detected using observation platform, according to the geometries characteristic information for being observed platform, are established space and are crossed moment forecasting procedure；It establishes based on the geometries characteristic information, the Kalman filter of gray feature information and geometric area characteristic information or the calculation method of ratio characteristic for being observed platform；Observation platform is forecast constantly with crossing for platform is observed；The accurate forecast for being observed platform gray feature information, geometries characteristic information and the metastable feature of geometric area characteristic information, obtaining to the moment that crosses that this method is obtained using passive measurement；More traditional passive ranging method does not need crossrange maneuvering process, while more traditional passive ranging method has higher convergent probability and convergence precision.
Description
Technical field
The invention belongs to utilize radio wave ranging technology field, and in particular to a kind of crossing the moment based on target signature
Forecasting procedure.
Background technique
During platform is close to each other, to the accurate forecast at the moment that crosses in Guidance and control, platform anticollision, cross pair
Connecing operation etc. has an important Practical significance, and tradition crosses forecast there are two main classes the method at moment, and one kind is using actively
Distance measuring method measures range and range rate platform, then calculates the time information that crosses, and has strong real-time, essence
Spend high advantage, but need to increase complicated initiative range measurement system, it is difficult to adapt to it is remote (dozens of kilometres to several hundred kilometers away from
From) process that quickly crosses of target.One kind is to pass through filtering using the observation angle information of target using passive ranging method
Method is adjusted the distance is settled accounts with velocity information, but is needed observation platform and be observed between platform the transverse direction for having a certain distance
Mobile process could eliminate more solution problems present in filtering, obtain more accurate distance and velocity information, crossrange maneuvering
Additional power and time of measuring are needed, is very restricted in the sky with space application.
Summary of the invention
For the above-mentioned prior art, the purpose of the present invention is to provide a kind of forecast side constantly that crosses based on target signature
Method, the gray feature information for being observed platform, geometries characteristic information and the geometric area that can be obtained using passive measurement
The metastable feature of characteristic information is not needing observation platform and is being observed crossrange maneuvering and other Distance-sensings between platform
In the case where device, accurate forecast is carried out constantly to crossing.
In order to achieve the above object, the present invention uses following technical scheme.
A kind of moment forecasting procedure that crosses based on target signature of the present invention, method includes the following steps:
(1) observation platform detects the geometries characteristic information, gray feature information and geometric area for being observed platform
Characteristic information, the geometries characteristic information for being observed platform provided according to observation platform imaging detection system, establishes space
Cross moment forecasting procedure；
When the geometric dimension for being observed platform imaging in observation platform detection system is less than or equal to 3 pixel, execute
Step (2)；
When the geometric dimension for being observed platform imaging in observation platform detection system is greater than 3 pixel, step is executed
(3)；
(2) it establishes based on the moment forecasting procedure that crosses for being observed platform gray feature, includes the following steps:
Step 1: establishing the analytical relation for being observed platform gray feature with the moment that crosses, and derives and is observed platform ash
The parsing relationship of degree feature and the remaining time that crosses；
Step 2: judgement is observed the signal to noise ratio of platform gray feature in observation platform detection system, when being observed platform
When the signal to noise ratio of gray feature in observation platform detection system is less than or equal to 20, step 3 in step (2) is executed；When being observed
Platform executes step 4 in step (2) when the signal to noise ratio of gray feature in observation platform detection system is greater than 20；
Step 3: according to the analytical relation being derived by step 1 in step (2), initial proportion coefficient, residue are selected
Time, the residue of the crossing time-derivative that crosses are quantity of state, and being observed platform gray feature is observed quantity, is established flat based on being observed
The Kalman filter of platform gray feature；
Step 4: the solution for being observed platform gray feature and the remaining time that crosses established according to step 1 in step (2)
Analysis relationship selects the different observation moment to be observed the gray feature of platform, establishes residue and crosses and the time and is observed platform gray scale
The calculation method of the ratio characteristic of feature；
Step 5: forecasting observation platform with crossing for platform is observed constantly, and output residue, which crosses, the time and to cross
Moment predicted value；
(3) it establishes based on the moment forecasting procedure that crosses for being observed platform geometric area feature or geometries characteristic, packet
Include following steps:
Step 1: establishing the analytical relation for being observed the moment that crosses of platform geometric area feature or geometries characteristic,
Derive the parsing relationship for being observed platform geometric area feature or geometries characteristic and the remaining time that crosses；
Step 2: judgement is observed platform geometric area feature or geometries characteristic in observation platform detection system
Signal to noise ratio is less than when being observed platform signal to noise ratio of geometric area feature or geometries characteristic in observation platform detection system
When equal to 20, step 3 in step (3) is executed；When being observed platform geometric area feature or several in observation platform detection system
When the signal to noise ratio of what size characteristic is greater than 20, step 4 in step (3) is executed；
Step 3: according to the analytical relation being derived by step 1 in step (3), selection is observed platform geometric area
Feature or geometries characteristic, residue time, the residue time-derivative that crosses that crosses are quantity of state, and it is special to be observed platform geometric area
Sign or geometries characteristic are observed quantity, establish the Kalman filtering for being observed platform geometric area feature or geometries characteristic
Device；
Step 4: platform geometric area feature or geometries characteristic are observed according to what step 1 in step (3) was established
With the parsing relationship of the remaining time that crosses, the different observation moment is selected to be observed geometric area feature or the geometric dimension spy of platform
Sign establishes the resolving side of remaining cross time and the ratio characteristic for being observed platform geometric area feature or geometries characteristic
Method；
Step 5: forecasting observation platform with crossing for platform is observed constantly, and output residue, which crosses, the time and to cross
Moment predicted value.
Step 1 is observed the analytical relation of platform gray feature with the moment that crosses in the step (2) are as follows:
In formula: T indicates that residue crosses the time, and unit is s；gTar(T) gray feature of platform is observed for the T moment；R is
It is observed distance of the platform between observation platform optical imaging system, unit is m；K is observation platform detection system conversion system
Number；ITarTo be observed platform radiation intensity；V is observation platform and is observed closing speed between platform, and unit is m/s；For observation platform detection system conversion coefficient K, it is observed platform radiation intensity ITarWith observation platform and be observed platform
Between v squares of closing speed of ratio, estimated according to observation platform detection result.
Step 1 is observed the analytical relation of platform geometric area feature and the remaining time that crosses in the step (3)
Are as follows:
In formula: SsenIt (T) is the geometric area feature for being observed platform the T moment in observed image, unit is pixel；SObj
It is the geometric area feature for being observed platform, unit is pixel；F is optical imaging system focal length, unit m；V is observation platform
Be observed closing speed between platform, unit m/s；SPixIt is observation platform detection system Pixel Dimensions, unit m；T is
Residue crosses the time, unit s；
Step 1 is observed the analytical relation of platform geometries characteristic and the remaining time that crosses in the step (3)
Are as follows:
In formula: LsenIt (T) is to be observed geometries characteristic of the platform in observed image the T moment, unit is pixel；LObj
It is the geometries characteristic for being observed platform, unit is pixel；F is optical imaging system focal length, unit m；V is observation platform
Be observed closing speed between platform, unit m/s；SPixIt is observation platform detection system Pixel Dimensions, unit m；T is
Residue crosses the time, unit s.
Kalman filter relational expression is equal in step 3 in step 3 and step (3) in the step (2) are as follows:
In formula: X (k+1) is value of the quantity of state X at the k+1 moment；X (k) is value of the quantity of state X at the k moment；Δ t is the time
Interval, unit s；Φ is state-transition matrix.
Crossing for Kalman filter forecasts iteration constantly in step 3 in step 3 and step (3) in the step (2)
Calculating process are as follows:
A. predictive equation is calculated:
In formula:For state a step of forecasting value；Φ is state-transition matrix；XkFor vector X the k moment value；
B. it calculates mensuration and is generalized into the Jacobi matrix in future position:
In formula: Hk+1For the Jacobi matrix of k+1；G [X] is the right end nonlinear function vector of systematic observation equation group；X
For state variable；For state a step of forecasting value；
C. prediction covariance and Kalman gain are calculated:
In formula: PK+1/KFor a step of forecasting variance matrix；ΦK+1/KFor state-transition matrix；PK/KFor filtering error variance matrix；For matrix ΦK+1/KTransposed matrix；KK+1For filtering gain matrix；Hk+1For the Jacobi matrix of k+1；For matrix
Hk+1Transposed matrix；RK+1For systematic observation noise variance matrix；
D. filtering estimated value is obtained:
In formula:It is vector X according to the value at k+1 moment to the estimated value of k+1 moment value；For one step of state
Predicted value；KK+1For filtering gain matrix；GK+1For the observation vector at k+1 moment；For according to state a step of forecasting valueThe forecast vector of calculating；
E. prediction point data is reused if using Modified covariance extended Kalman filter device to recalculate
Jacobi matrix:
In formula:Jacobi matrix for the k+1 recalculated with prediction point data；G [X] is systematic observation equation group
Right end nonlinear function vector；X is state variable；It is that vector X estimates k+1 moment value according to the value at k+1 moment
Meter；
F. prediction covariance and Kalman gain are recalculated:
In formula:For the filtering gain matrix recalculated with prediction point data；For matrixTransposition square
Battle array；PK+1/KFor a step of forecasting variance matrix；PK+1/K+1For the correction value of K+1 moment variance matrix；Hk+1For the Jacobi square of k+1
Battle array；Jacobi matrix for the k+1 recalculated with prediction point data；For direction amountTransposed matrix；RK+1
For systematic observation noise variance matrix；I is unit vector matrix；T is that residue crosses the time, unit s；
G. it obtains can use after prediction result residue to cross time T forecast estimation filter value；
Step A to step F is repeated, is forecast constantly to crossing, residue is obtained and crosses and the time and cross moment predicted value.
In the establishment process of Kalman filter, the residue time-derivative that crosses is set as fixed value -1, and residue crosses the time
The standard variance of derivative is set as the value between 0.01~0.00001.
In the step (2) step 4 residue cross the time and be observed platform gray feature ratio characteristic resolving
Method relational expression are as follows:
In formula: T is that residue crosses the time, unit s；Δ t is time interval, unit s；gTarIt (T) is to be seen at the T moment
The gray value for surveying platform, correspondingly, gTar(T- Δ t) is the gray value that T- time Δt is observed platform.
Step 4 residue, which crosses, in the step (3) time and is observed the ratio characteristic of platform geometric area feature
Calculation method relational expression are as follows:
In formula: T is that residue crosses the time, unit s；Δ t is time interval, unit s；STarIt (T) is to be seen at the T moment
The two dimensional image geometric area feature that platform is presented in observation platform detection system is surveyed, unit is pixel；Correspondingly, STar(T-
Δ t) is the two dimensional image geometric area feature that T- time Δt is observed that platform is presented in observation platform detection system, unit
It is pixel；
Step 4 residue, which crosses, in the step (3) time and is observed the ratio characteristic of platform geometries characteristic
Resolve relational expression are as follows:
In formula: T is that residue crosses the time, unit s；Δ t is time interval, unit s；LTarIt (T) is to be seen at the T moment
The one dimensional image geometries characteristic that platform is presented in observation platform detection system is surveyed, unit is pixel；Correspondingly, LTar(T-
Δ t) is the one dimensional image geometries characteristic that T- time Δt is observed that platform is presented in observation platform detection system, unit
It is pixel.
The Kalman filter is Modified covariance extended Kalman filter device.
Technical solution provided in an embodiment of the present invention has the benefit that
A kind of moment forecasting procedure that crosses based on target signature of the present invention, carries out constantly to close to crossing between platform
Accurate forecast, this method are observed platform gray feature information, geometries characteristic information and several using what passive measurement obtained
What metastable feature of area features information obtains the accurate forecast to the moment that crosses；More traditional passive ranging method, is not required to
Crossrange maneuvering process is wanted, while more traditional passive ranging method has higher convergent probability and convergence precision.
Detailed description of the invention
Fig. 1 is the moment forecasting procedure flow chart that crosses the present invention is based on target signature；
Fig. 2 is to be estimated based on the platform grayscale emulational sequence that is observed for being observed platform gray feature with Kalman filter
Gray value curve graph；
Fig. 3 be crossed based on the Kalman filter for being observed platform gray feature moment forecasting procedure direct filtering it is pre-
Report value and true value curve graph；
Fig. 4 be crossed based on the Kalman filter for being observed platform gray feature moment forecasting procedure absolute error it is bent
Line chart；
Fig. 5 be crossed based on the Kalman filter for being observed platform gray feature moment forecasting procedure relative error it is bent
Line chart；
When Fig. 6 is time interval Δ t=2.5s, platform is observed based on what ratio resolved that relationship crosses moment forecasting procedure
Grayscale emulational sequence curve figure；
When Fig. 7 is time interval Δ t=2.5s, relationship is resolved based on the ratio for being observed platform gray feature and is crossed the moment
The predicted value and true value curve graph of forecasting procedure；
When Fig. 8 is time interval Δ t=2.5s, relationship is resolved based on the ratio for being observed platform gray feature and is crossed the moment
The absolute error curve graph of forecasting procedure；
When Fig. 9 is time interval Δ t=2.5s, relationship is resolved based on the ratio for being observed platform gray feature and is crossed the moment
The relative error curve graph of forecasting procedure；
Figure 10 is that being observed for moment forecasting procedure that crossed based on the Kalman filter for being observed platform geometrical characteristic is flat
Table top accumulates characteristics simulation sequence curve figure；
Figure 11 is the direct filtering of moment forecasting procedure of being crossed based on the Kalman filter for being observed platform geometrical characteristic
Predicted value and true value curve graph；
Figure 12 is the absolute error of moment forecasting procedure of being crossed based on the Kalman filter for being observed platform geometrical characteristic
Curve graph；
When Figure 13 is time interval Δ t=2.5s, when being crossed based on the ratio resolving relationship for being observed platform geometrical characteristic
Carve the predicted value and true value curve graph of forecasting procedure；
When Figure 14 is time interval Δ t=2.5s, when being crossed based on the ratio resolving relationship for being observed platform geometrical characteristic
Carve the absolute error curve graph of forecasting procedure.
Specific embodiment
It crosses the moment to a kind of space based on target observation feature of the present invention with reference to the accompanying drawings and detailed description
Forecasting procedure elaborates.
Moment forecasting procedure, including following step as shown in Figure 1, a kind of space based on target observation feature of the present invention crosses
It is rapid:
Observation platform imaging detection system provides the gray feature information for being observed platform, geometries characteristic information and several
What area features information establishes the moment forecasting procedure that crosses according to the geometries characteristic information for being observed platform；
One, it when the geometric dimension for being observed platform imaging in observation platform detection system is less than or equal to 3 pixel, builds
It is based on and is observed the moment forecasting procedure that crosses of platform gray feature, include the following steps:
(1) analytical relation for being observed platform gray feature with the moment that crosses is established, derives and is observed platform gray scale
The parsing relationship of feature and the remaining time that crosses；
According to the imaging process of observation platform passive measurement system, the radiation of platform, background is observed by optical imagery
System is converted to digital picture, is observed platform, the radiation intensity of background is presented as gray scale in target digital image.By seeing
After surveying plateform system correction, calibration, 1 table of formula can be used by being observed the relationship between the radiation feature of platform and gray feature
Show.
In formula: g(Bac+Tar)To be observed the gray scale that platform and background radiation are formed on observation platform optical imaging system；
ITarFor the radiation intensity for being observed platform, unit is W/Sr；IBacFor the radiation intensity of background, unit is W/Sr；R is to be observed
Distance of the platform between observation platform optical imaging system, unit are m；K is that observation platform optical imaging system is strong to radiating
The conversion proportion coefficient of image grayscale is spent, unit is m2Sr/W；B is observation platform optical imaging system background image gray scale；
gTarTo be observed the gray scale that platform is formed on observation platform optical imaging system；gBacIt is background radiation in observation platform optics
The gray scale formed in imaging system.
Indicate that residue crosses the time with T, unit is s, and the residue time T that crosses is gradually reduced；V be observation platform be observed
Closing speed between platform, unit are m/s；Then formula 1 can be rewritten as:
In formula: T indicates that residue crosses the time, and unit is s；gTar(T) gray feature of platform is observed for the T moment；R is
It is observed distance of the platform between observation platform optical imaging system, unit is m；K is observation platform detection system conversion system
Number；ITarTo be observed platform radiation intensity；V is observation platform and is observed closing speed between platform, and unit is m/s；For observation platform detection system conversion coefficient K, it is observed platform radiation intensity ITarWith observation platform and be observed platform
Between v squares of closing speed of ratio, estimated according to observation platform detection result.
Due in formula (2)It is observation platform detection system conversion coefficient K, is observed platform radiation intensity ITar
With observation platform and be observed v squares of closing speed between platform of ratio, according to close hypothesis condition and be observed platform
The metastable feature of radiation feature, observation platform detection system conversion coefficient K are observed platform radiation intensity ITarAnd observation
Platform and to be observed closing speed v between platform be definite value；It is therefore contemplated that when being observed platform gray feature and only crossing with residue
Between T it is related, and at inverse square relationship.
From the foregoing, it will be observed that using passive measurement to be observed platform gray feature, can accurately estimate the moment that crosses, no
Other operations such as need additional lateral motor-driven.
(2) judgement is observed the signal to noise ratio of platform gray feature in observation platform detection system, exists when being observed platform
It is special based on (1) step observation platform gray scale in step 1 when the signal to noise ratio of gray feature is less than 20 in observation platform detection system
The analytical relation of sign and the remaining time that crosses, selection are observed platform initial gray GGray0, residue cross time T, remaining hand over
Remittance time-derivative T' is state variable, is observed the gray feature g of platformTar(T) it is observed quantity, establishes Kalman filter.
Using formula 2, state variable is selected to select X=(GGray0T T') ', wherein GGray0It is handed over for initial gray, T residue
It converges the time, T' is the first derivative of T.Ideally, state equationObservational equationThe foundation of specific Kalman filter filtering is as follows:
X=(GGray0T T') '=(G0 T T')TFormula 3
After discretization are as follows:
In formula: GGray0To be observed platform initial gray value, unit is pixel；T residue crosses the time, unit s；T' is
The first derivative of T；V (T) is ambient noise；X (k+1) is value of the quantity of state X at the k+1 moment；X (k) is quantity of state X at the k moment
Value；Δ t is time interval, unit s；Φ is state-transition matrix.
Due to observational equationContaining nonlinear terms, therefore Kalman filter uses extension
Kalman filter (EKF) can also use Modified covariance extended Kalman filter to increase the estimation to nonlinear terms
Device (MCEKF) can improve the case where common EKF filter is to convergence.
Due to the presence of the nonlinear terms of EKF and MCEKF, when initial estimation parameter setting is unreasonable, it will cause to filter
The convergence rate of wave device is slow, therefore parameter setting has large effect to filter effect, is especially observed the initial ash of platform
Angle value GGray0, residue cross time T, be observed platform initial gray value GGray0Square estimated value E [G0 2], residue is when crossing
Between T squares of initial value estimated value E [T0 2] initial value setting, the convergence rate and the essence after convergence for influencing Kalman filter
Degree, when being observed platform initial gray value GGray0Square estimated value E [G0 2], the estimated value E [T of T squares of initial value of time0 2]
When parameter setting is larger, fast convergence rate, but shaken greatly after restraining, filtering accuracy is lower；When being observed platform initial gray value
GGray0Square estimated value E [G0 2], the estimated value E [T of T squares of initial value of time0 2] parameter setting is too small, convergence rate by
It influences, but restrains post filtering value and stablize, precision is high.
Based on the Design on Kalman Filter method of passive measurement process in the design of filter, do not occur close to speed
It spends, be observed platform radiation intensity (being observed platform geometric area feature or geometries characteristic), observation platform and be observed
The information such as the distance between platform, from formula 6 as can be seen that only there are 3 quantity of states in established Kalman filter,
Such filter design structure letter, variable is few, can significantly enhance the convergent probability and convergence rate of filter.
In the establishment process of Kalman filter, in the state equation of selection, the residue time T that crosses is gradually decreased,
The time first derivative T' that residue can be crossed is set as fixed value -1, and the number of unknown quantity, mentions in this way in reduction state quantity space
The high convergent probability of filter, while its standard variance is set as the value between 0.01~0.00001, in acceptable filtering mistake
In difference, the convergence rate of filter is improved.
(3) Kalman filter established based on (2) step in step 1, crosses to observation platform with platform is observed
Moment forecast, is obtained residue and is crossed time T and the moment predicted value that crosses；
The forecast iterative process constantly that crosses based on Kalman filter is as follows:
A. predictive equation is calculated:
In formula:For state a step of forecasting value；Φ is state-transition matrix；XkFor vector X the k moment value；
B. it calculates mensuration and is generalized into Jacobi (Jacobi) matrix in future position:
In formula: Hk+1For Jacobi (Jacobi) matrix of k+1；G [X] is the non-linear letter of right end of systematic observation equation group
Number vector；X is state variable；For state a step of forecasting value；
C. prediction covariance and Kalman (Kalman) gain are calculated:
In formula: PK+1/KFor a step of forecasting variance matrix；ΦK+1/KFor state-transition matrix；PK/KFor filtering error variance matrix；For matrix ΦK+1/KTransposed matrix；KK+1For filtering gain matrix；Hk+1For Jacobi (Jacobi) matrix of k+1；For matrix Hk+1Transposed matrix；RK+1For systematic observation noise variance matrix；
D. filtering estimated value is obtained:
In formula:It is vector X according to the value at k+1 moment to the estimated value of k+1 moment value；For one step of state
Predicted value；KK+1For filtering gain matrix；GK+1For the observation vector at k+1 moment；For according to state a step of forecasting valueThe forecast vector of calculating
E. prediction point data is reused if using Modified covariance extended Kalman filter device (MCEKF) to count again
Calculate Jacobi (Jacobi) matrix:
In formula:Jacobi (Jacobi) matrix for the k+1 recalculated with prediction point data；G [X] is systematic perspective
Survey the right end nonlinear function vector of equation group；X is state variable；Be vector X according to the value at k+1 moment to k+1 when
Quarter value estimation；
F. prediction covariance and Kalman (Kalman) gain are recalculated:
In formula:For the filtering gain matrix recalculated with prediction point data；For matrixTransposition square
Battle array；PK+1/KFor a step of forecasting variance matrix；PK+1/K+1For the correction value of K+1 moment variance matrix；Hk+1For the Jacobi of k+1
(Jacobi) matrix；Jacobi (Jacobi) matrix for the k+1 recalculated with prediction point data；For direction amountTransposed matrix；RK+1For systematic observation noise variance matrix；I is unit vector matrix；T is that residue crosses the time, unit
s；
G. it obtains can use after prediction result residue to cross time T forecast estimation filter value；
Step A to step F is repeated, is forecast constantly to crossing, residue is obtained and crosses and the time and cross moment predicted value.
(4) when the signal to noise ratio for being observed platform gray feature in observation platform detection system is greater than 20, according to step
What (1) step was established in one is observed the parsing relationship of platform gray feature and the remaining time that crosses, when selecting different observations
The gray feature for being observed platform is carved, the resolving side of remaining cross time and the ratio characteristic for being observed platform gray feature are established
Method；
It is observed platform gray feature sequence based on what is obtained in observation platform detection system, selects time interval for Δ t
Two observation moment be observed the gray value g of platformTar(T)、gTar(T- Δ t), available according to formula 3:
In formula: T is that residue crosses the time, unit s；Δ t is time interval, unit s；gTarIt (T) is to be seen at the T moment
The gray value for surveying platform, correspondingly, gTar(T- Δ t) is the gray value that T- time Δt is observed platform.
The parsing relationship for being observed platform gray feature and the remaining time that crosses established using formula 15, can be direct
Residue is calculated to cross the time.
It is above-mentioned to be set based on the Kalman filter forecasting process initial parameter for being observed platform gray feature are as follows:
It is assumed that under vacuum conditions, observation platform and it is observed initial distance 13000m between platform, it is at the uniform velocity close between platform,
Closing speed is 1000m/s；Being observed platform initially is on the detection system point target, is observed the radiation intensity feature of platform
It keeps stablizing, to observation platform and is observed crossing between platform according to above-mentioned condition and forecasts constantly；It is wherein known to see
Survey platform maximum detectable range 10000m, observation platform and the typical closing speed being observed between platform be 700m/s~
1200m/s.Assuming that optical imaging system Space Angle 0.05mrad, is observed platform size in 1m~10m.
Initial predicted time value is set as 16, since the time is fixed reduction, so being T'=-1；Gray scale initial estimate
Setting 100, then filter original stateIn formula,For filter error variance matrix setup values.
Systematic observation error statistics feature-set is as follows: the estimated value E [V of observation noise V initial value0]=0, observation noise
Estimated value E [the V of V squares of initial valueK 2]=50.
Initial value observational characteristic sets as follows according to observing capacity: the estimated value E [G of gray value G initial value0]=0, gray scale
Estimated value E [the G of G squares of initial value of value0 2]=5 × 104；Estimated value E [the T of time T initial value0]=0, T squares of the time initial
Estimated value E [the T of value0 2]=25；The first derivative estimated value E [T' of time T0]=0, the first derivative square of time T it is initial
It is worth estimated value E [T0'2]=0.0001.Work as T'0It is fixed value -1, belongs to known quantity, works as T'0Deviate -1, filter result will occur
E [T is arranged in order to guarantee the correct convergence of filter in mistake0'2]=0.0001, it is ensured that T'0With very small deviation profile-
Near 1.
Fig. 2 to Fig. 5 is the emulation of moment forecasting procedure of being crossed based on the Kalman filter for being observed platform gray feature
Forecast result schematic diagram.It is can be seen that from Fig. 2 to Fig. 5 according to simulation analysis, based on the filter that the above method is established, forecast
Error is less than 10%, in the range of filter converges to prediction error less than 10% in 4~6s.
When Fig. 6 to Fig. 9 gives Δ t=2.5s, when being crossed based on the ratio resolving relationship for being observed platform gray feature
Carve the emulation forecast result schematic diagram of forecasting procedure.It is can be seen that from Fig. 6 to Fig. 9 when signal to noise ratio is higher, is observed platform gray scale
When changing features are very fast, prediction error is less than 10%.
Two, when the geometric dimension for being observed platform imaging in observation platform detection system is greater than 3 pixel, base is established
In the moment forecasting procedure that crosses for being observed platform geometrical characteristic, include the following steps:
(1) analytical relation for being observed platform geometrical characteristic with the moment that crosses is established, derives and is observed platform geometry
The parsing relationship of feature and the remaining time that crosses；
It is observed platform in the imaging session for being observed platform according to observation platform optical system imaging relationship and is observing
The size of Pixel Dimensions, expression formula are as follows in platform detection system:
In formula, LsenIt is the pixel size for being observed platform in observation platform detection system, unit is pixel；R target away from
With a distance from detection system, unit is m；F is the focal length of observation platform optical imagery detection system, and unit is m；LObjIt is to be seen
The one-dimensional geometric dimension of platform is surveyed, unit is m；SPixIt is the Pixel Dimensions of observation platform detection system, unit is pixel.
With the derivation of formula 2, indicate that residue crosses the time with T, unit is s, and the residue time T that crosses is gradually reduced；V is to see
It surveys platform and is observed closing speed between platform, unit is m/s；It can be derived by from formula 17:
It can then derive the geometric area feature for being observed platform in observation platform detection system:
The forecast of the remaining time that crosses is carried out with the geometric area feature for being observed platform using formula 18, it equally can also be with
Quarter is handed over the one-dimensional geometries characteristic forecast for being observed platform according to formula 17；Utilize the geometrical characteristic information for being observed platform
Forecast crosses the moment, it is not necessary to add any crossrange maneuvering process.
(2) it is miscellaneous to be observed platform letter of geometric area feature or geometries characteristic in observation platform detection system for judgement
Than, when be observed platform in observation platform detection system the signal to noise ratio of geometric area feature or geometries characteristic less than 20
When, the analytical relation that can be derived according to (1) step in step 2, selection is observed the initial geometric surface product value S of platformsen0Or just
Beginning geometric dimension value Lsen0, residue cross time T, residue cross time-derivative T' be state variable, be observed the geometry of platform
Area features Ssen(T) or geometries characteristic Lsen(T) it is observed quantity, establishes Kalman filter.
Below to be observed the geometric area feature S of platformsen(T) to illustrate building for Kalman filter for observed quantity
Vertical process, is observed platform geometries characteristic Lsen(T) for observed quantity establish Kalman filter process it is prefabricated similar.
Using formula 18, state variable is selected to select X=(Ssen0T T') ', wherein Ssen0For initial geometric surface product value, T
Residue crosses the time, and T' is the first derivative of T.Ideally, state equationObservational equationThe residue time T that crosses is predicted and estimated using Kalman filter.
The foundation of specific Kalman filter filtering is as follows:
X=(Ssen0T T') '=(Ssen0 T T')TFormula 19
After discretization are as follows:
In formula: Ssen0For initial geometric surface product value, unit is pixel；T residue crosses the time, unit s；T' is the one of T
Order derivative；V (T) is ambient noise；X (k+1) is value of the quantity of state X at the k+1 moment；X (k) is value of the quantity of state X at the k moment；
Δ t is time interval, unit s；Φ is state-transition matrix.
Due to observational equationContaining nonlinear terms, therefore Kalman filter uses expansion
It opens up Kalman filter (EKF), in order to increase to the non-linear estimation thought, can also be filtered using modified covariance method spreading kalman
Wave device (MCEKF) can improve the case where common EKF filter is to convergence.
Based on the Design on Kalman Filter method of passive measurement process in the design of filter, do not occur close to speed
Spend, be observed platform radiation intensity (being observed platform geometrical characteristic), observation platform and be observed the letter such as the distance between platform
Breath, from formula 22 as can be seen that only occurring 3 quantity of states in established Kalman filter, such filter design
Structure letter, variable is few, can significantly enhance the convergent probability and convergence rate of filter.
In the establishment process of Kalman filter, in the state equation of selection, the residue time T that crosses is gradually decreased,
The time first derivative T' that residue can be crossed is set as fixed value -1, and the number of unknown quantity, mentions in this way in reduction state quantity space
The high convergent probability of filter, while its standard variance is set as the value between 0.01~0.00001, in acceptable filtering mistake
In difference, the convergence rate of filter is improved.
(3) it based on the Kalman filter established in step 2 (2) step, crosses to observation platform with platform is observed
Moment forecast, is obtained residue and is crossed time T and the moment predicted value that crosses；The iterative process and step of Kalman filter
(3) step iterative process A to F is consistent in rapid one.
(4) it is miscellaneous to be observed platform letter of geometric area feature or geometries characteristic in observation platform detection system for judgement
Than being greater than 20 when being observed platform signal to noise ratio of geometric area feature or geometries characteristic in observation platform detection system
When, the parsing relationship for being observed platform geometrical characteristic and the remaining time that crosses established according to (1) step in step 2, selection
The different observation moment are observed the geometrical characteristic of platform, establish the remaining ratio spy for crossing the time and being observed platform geometrical characteristic
The calculation method of sign；
It is observed platform geometric area feature or geometries characteristic sequence based on what is obtained in observation platform detection system,
Time interval is selected to be observed the geometrical characteristic of platform for the two observation moment of Δ t, available residue, which crosses, the time and to be seen
Survey the resolving relational expression of the ratio characteristic of platform geometric area feature are as follows:
In formula: T is that residue crosses the time, unit s；Δ t is time interval, unit s；STarIt (T) is to be seen at the T moment
The two dimensional image geometric area feature that platform is presented in observation platform detection system is surveyed, unit is pixel；Correspondingly, STar(T-
Δ t) is the two dimensional image geometric area feature that T- time Δt is observed that platform is presented in observation platform detection system, unit
It is pixel.
With formula 23, the resolving relational expression of remaining cross time and the ratio characteristic for being observed platform geometries characteristic
Are as follows:
In formula: T is that residue crosses the time, unit s；Δ t is time interval, unit s；LTarIt (T) is to be seen at the T moment
The one dimensional image geometries characteristic that platform is presented in observation platform detection system is surveyed, unit is pixel；Correspondingly, LTar(T-
Δ t) is the one dimensional image geometries characteristic that T- time Δt is observed that platform is presented in observation platform detection system, unit
It is pixel.
Platform geometric area feature or geometries characteristic and residue are observed using what formula 23 or formula 24 were established
Cross the parsing relationship of time, can go out residue with directly calculation and cross the time.
It is above-mentioned based at the beginning of the Kalman filter forecasting process for being observed platform geometric area feature or geometries characteristic
Value parameter setting are as follows:
It is assumed that under vacuum conditions, observation platform and it is observed initial distance 130m between platform, it is at the uniform velocity close between platform, it connects
Nearly speed is 10m/s；Being observed platform initially is on the detector spot target, and the radiation intensity feature of target keeps stablizing, root
The intersection moment platform is forecast according to above-mentioned condition；Wherein known passive measurement platform maximum detectable range 200m is put down
Typical closing speed between platform is 7m/s~15m/s.Assuming that optical imaging system Space Angle 1.0mrad, target size is in 0.5m
~3m.
Systematic observation error statistics feature-set is as follows: the estimated value E [V of observation noise V initial value0]=0, observation noise
Estimated value E [the V of V squares of initial valueK 2]=1.
Initial value observational characteristic sets as follows according to observing capacity: maximum detectable range 200m, and closing speed about 7~
15m/s, therefore initial predicted time value is set as 18s, since the time is fixed reduction, so being T'=-1；Gray scale is initially estimated
Evaluation setting 20, then filter original stateSystematic observation error statistics feature E [V0]=0, E
[VK 2]=10；Initial value observational characteristic sets following E [G according to observing capacity0]=0, E [G0 2]=2 × 102；E[T0]=0, E
[T0 2]=25；E[T'0]=0, works as T'0Deviateing -1, mistake will occur in filter result, in order to guarantee the correct convergence of filter, if
Set E [T0'2]=0.0001, it is ensured that T'0With very small deviation profile near -1.
Figure 10 to Figure 12 is given to be crossed based on the Kalman filter for being observed platform area geometrical characteristic and be forecast constantly
The emulation forecast result schematic diagram of method, straight line is true value in Figure 11, and the curve of cyclical fluctuations is Kalman filtering value.Figure 13 and Figure 14
When giving Δ t=2.5s, relationship is resolved based on the ratio for being observed platform area geometrical characteristic and is crossed moment forecasting procedure
Forecast result schematic diagram is emulated, straight line is true value in Figure 13, and the curve of cyclical fluctuations is Kalman filtering value.
The moment forecasting procedure that crosses from step 1 and step 2 can be seen that using being observed platform gray feature, several
What area features or geometries characteristic and the relationship of the remaining time that crosses are crossed and are forecast constantly, not seen completely with additional
The crossrange maneuvering or other supplementary observation processes for surveying platform device just can be realized the forecast to the remaining time that crosses.
Claims (9)
1. a kind of moment forecasting procedure that crosses based on target signature, it is characterised in that: the following steps are included:
(1) observation platform detects the geometries characteristic information, gray feature information and geometric area feature for being observed platform
Information, the geometries characteristic information for being observed platform provided according to observation platform imaging detection system, establishes space and crosses
Moment forecasting procedure；
When the geometric dimension for being observed platform imaging in observation platform detection system is less than or equal to 3 pixel, step is executed
(2)；
When the geometric dimension for being observed platform imaging in observation platform detection system is greater than 3 pixel, execute step (3)；
(2) it establishes based on the moment forecasting procedure that crosses for being observed platform gray feature, includes the following steps:
Step 1: establishing the analytical relation for being observed platform gray feature with the moment that crosses, and derives and is observed platform gray scale spy
The parsing relationship of sign and the remaining time that crosses；
Step 2: judgement is observed the signal to noise ratio of platform gray feature in observation platform detection system, is seeing when being observed platform
When surveying the signal to noise ratio of gray feature in platform detection system less than or equal to 20, step 3 in step (2) is executed；When being observed platform
When the signal to noise ratio of gray feature in observation platform detection system is greater than 20, step 4 in step (2) is executed；
Step 3: according to the analytical relation being derived by step 1 in step (2), initial proportion coefficient, residue is selected to cross
Time, residue cross time-derivative for quantity of state, and being observed platform gray feature is observed quantity, establish grey based on platform is observed
Spend the Kalman filter of feature；
Step 4: the parsing pass for being observed platform gray feature and the remaining time that crosses established according to step 1 in step (2)
System selects the different observation moment to be observed the gray feature of platform, establishes residue and crosses and the time and is observed platform gray feature
Ratio characteristic calculation method；
Step 5: forecasting observation platform with crossing for platform is observed constantly, and output residue, which crosses, the time and to cross the moment
Predicted value；
(3) it establishes based on the moment forecasting procedure that crosses for being observed platform geometric area feature or geometries characteristic, including such as
Lower step:
Step 1: establishing the analytical relation for being observed platform geometric area feature or geometries characteristic with the moment that crosses, and derives
It is observed the parsing relationship of platform geometric area feature or geometries characteristic and the remaining time that crosses out；
Step 2: it is miscellaneous that judgement is observed platform letter of geometric area feature or geometries characteristic in observation platform detection system
Than being less than or equal to when being observed platform signal to noise ratio of geometric area feature or geometries characteristic in observation platform detection system
When 20, step 3 in step (3) is executed；When being observed platform geometric area feature or dimensioning in observation platform detection system
When the signal to noise ratio of very little feature is greater than 20, step 4 in step (3) is executed；
Step 3: according to the analytical relation being derived by step 1 in step (3), selection is observed platform geometric area feature
Or geometries characteristic, residue time, the residue time-derivative that crosses that crosses are quantity of state, be observed platform geometric area feature or
Geometries characteristic is observed quantity, establishes the Kalman filter for being observed platform geometric area feature or geometries characteristic；
Step 4: platform geometric area feature or geometries characteristic are observed and is remained according to what step 1 in step (3) was established
The parsing relationship of the remaining time that crosses selects the different observation moment to be observed the geometric area feature or geometries characteristic of platform,
Establish the calculation method of remaining cross time and the ratio characteristic for being observed platform geometric area feature or geometries characteristic；
Step 5: forecasting observation platform with crossing for platform is observed constantly, and output residue, which crosses, the time and to cross the moment
Predicted value.
2. the moment forecasting procedure that crosses according to claim 1 based on target signature, it is characterised in that: the step
(2) step 1 is observed the analytical relation of platform gray feature and the remaining time that crosses in are as follows:
In formula: T indicates that residue crosses the time, and unit is s；gTar(T) gray feature of platform is observed for the T moment；R is to be seen
Distance of the platform between observation platform optical imaging system is surveyed, unit is m；K is observation platform detection system conversion coefficient；
ITarTo be observed platform radiation intensity；V is observation platform and is observed closing speed between platform, and unit is m/s；
For observation platform detection system conversion coefficient K, it is observed platform radiation intensity ITarWith observation platform and be observed platform it is indirectly
The ratio of nearly v squares of speed, is estimated according to observation platform detection result.
3. the moment forecasting procedure that crosses according to claim 1 based on target signature, it is characterised in that: the step
(3) step 1 is observed the analytical relation of platform geometric area feature and the remaining time that crosses in are as follows:
In formula: SsenIt (T) is the geometric area feature for being observed platform the T moment in observed image, unit is pixel；SObjBe by
The geometric area feature of observation platform, unit are pixel；F is optical imaging system focal length, unit m；V is observation platform and quilt
Closing speed between observation platform, unit m/s；SPixIt is observation platform detection system Pixel Dimensions, unit m；T is remaining
It crosses the time, unit s；
Step 1 is observed the analytical relation of platform geometries characteristic and the remaining time that crosses in the step (3) are as follows:
In formula: LsenIt (T) is to be observed geometries characteristic of the platform in observed image the T moment, unit is pixel；LObjBe by
The geometries characteristic of observation platform, unit are pixels；F is optical imaging system focal length, unit m；V is observation platform and quilt
Closing speed between observation platform, unit m/s；SPixIt is observation platform detection system Pixel Dimensions, unit m；T is remaining
It crosses the time, unit s.
4. the moment forecasting procedure that crosses according to claim 2 or 3 based on target signature, it is characterised in that: described
Kalman filter relational expression is equal in step 3 in step 3 and step (3) in step (2) are as follows:
In formula: X (k+1) is value of the quantity of state X at the k+1 moment；X (k) is value of the quantity of state X at the k moment；Δ t is time interval,
Unit is s；Φ is state-transition matrix.
5. the moment forecasting procedure that crosses according to claim 4 based on target signature, it is characterised in that: Kalman filtering
Iterative process is forecast in crossing for device constantly are as follows:
A. predictive equation is calculated:
In formula:For state a step of forecasting value；Φ is state-transition matrix；XkFor vector X the k moment value；
B. it calculates mensuration and is generalized into the Jacobi matrix in future position:
In formula: Hk+1For the Jacobi matrix of k+1；G [X] is the right end nonlinear function vector of systematic observation equation group；X is shape
State variable；For state a step of forecasting value；
C. prediction covariance and Kalman gain are calculated:
In formula: PK+1/KFor a step of forecasting variance matrix；ΦK+1/KFor state-transition matrix；PK/KFor filtering error variance matrix；
For matrix ΦK+1/KTransposed matrix；KK+1For filtering gain matrix；Hk+1For the Jacobi matrix of k+1；For matrix Hk+1's
Transposed matrix；RK+1For systematic observation noise variance matrix；
D. filtering estimated value is obtained:
In formula:It is vector X according to the value at k+1 moment to the estimated value of k+1 moment value；For state a step of forecasting
Value；KK+1For filtering gain matrix；GK+1For the observation vector at k+1 moment；For according to state a step of forecasting value
The forecast vector of calculating；
E. prediction point data is reused if using Modified covariance extended Kalman filter device recalculates Jacobi square
Battle array:
In formula:Jacobi matrix for the k+1 recalculated with prediction point data；G [X] is the right side of systematic observation equation group
Terminal type non-linear functional vector；X is state variable；It is estimation of the vector X according to the value at k+1 moment to k+1 moment value；
F. prediction covariance and Kalman gain are recalculated:
In formula:For the filtering gain matrix recalculated with prediction point data；For matrixTransposed matrix；
PK+1/KFor a step of forecasting variance matrix；PK+1/K+1For the correction value of K+1 moment variance matrix；Hk+1For the Jacobi matrix of k+1；Jacobi matrix for the k+1 recalculated with prediction point data；For direction amountTransposed matrix；RK+1For
Systematic observation noise variance matrix；I is unit vector matrix；T is that residue crosses the time, unit s；
G. it obtains can use after prediction result residue to cross time T forecast estimation filter value；
Step A to step F is repeated, is forecast constantly to crossing, residue is obtained and crosses and the time and cross moment predicted value.
6. the moment forecasting procedure that crosses according to claim 4 based on target signature, it is characterised in that: filtered in Kalman
In the establishment process of wave device, the residue time-derivative that crosses is set as fixed value -1, and the cross standard variance of time-derivative of residue is set as
Value between 0.01~0.00001.
7. the moment forecasting procedure that crosses according to claim 2 based on target signature, it is characterised in that: the step
(2) in step 4 residue cross the time and be observed platform gray feature ratio characteristic calculation method relational expression are as follows:
In formula: T is that residue crosses the time, unit s；Δ t is time interval, unit s；gTar(T) to be that the T moment is observed flat
The gray value of platform, correspondingly, gTar(T- Δ t) is the gray value that T- time Δt is observed platform.
8. the moment forecasting procedure that crosses according to claim 3 based on target signature, it is characterised in that: the step
(3) in step 4 residue cross the time and be observed platform geometric area feature ratio characteristic calculation method relational expression are as follows:
In formula: T is that residue crosses the time, unit s；Δ t is time interval, unit s；STar(T) to be that the T moment is observed flat
The two dimensional image geometric area feature that platform is presented in observation platform detection system, unit is pixel；Correspondingly, STar(T-Δt)
It is the two dimensional image geometric area feature that T- time Δt is observed that platform is presented in observation platform detection system, unit is picture
Element；
In the step (3) step 4 residue cross the time and be observed platform geometries characteristic ratio characteristic resolving
Relational expression are as follows:
In formula: T is that residue crosses the time, unit s；Δ t is time interval, unit s；LTar(T) to be that the T moment is observed flat
The one dimensional image geometries characteristic that platform is presented in observation platform detection system, unit are pixels；Correspondingly, LTar(T-Δt)
It is the one dimensional image geometries characteristic that T- time Δt is observed that platform is presented in observation platform detection system, unit is picture
Element.
9. the moment forecasting procedure that crosses according to claim 1 based on target signature, it is characterised in that: the karr
Graceful filter is Modified covariance extended Kalman filter device.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410596963.2A CN105631063B (en) | 2014-10-30 | 2014-10-30 | A kind of moment forecasting procedure that crosses based on target signature |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410596963.2A CN105631063B (en) | 2014-10-30 | 2014-10-30 | A kind of moment forecasting procedure that crosses based on target signature |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105631063A CN105631063A (en) | 2016-06-01 |
| CN105631063B true CN105631063B (en) | 2019-01-08 |
Family
ID=56045995
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410596963.2A Active CN105631063B (en) | 2014-10-30 | 2014-10-30 | A kind of moment forecasting procedure that crosses based on target signature |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105631063B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112165368B (en) * | 2020-08-27 | 2021-12-21 | 西南科技大学 | Time-synchronized real-time adaptive convergence estimation system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1987355A (en) * | 2006-12-22 | 2007-06-27 | 北京航空航天大学 | Earth satellite self astronomical navigation method based on self adaptive expansion kalman filtering |
| CN102788976A (en) * | 2012-06-27 | 2012-11-21 | 北京理工大学 | High-order extended Kalman filtering method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5496008B2 (en) * | 2010-08-06 | 2014-05-21 | キヤノン株式会社 | Position / orientation measuring apparatus, position / orientation measuring method, and program |
-
2014
- 2014-10-30 CN CN201410596963.2A patent/CN105631063B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1987355A (en) * | 2006-12-22 | 2007-06-27 | 北京航空航天大学 | Earth satellite self astronomical navigation method based on self adaptive expansion kalman filtering |
| CN102788976A (en) * | 2012-06-27 | 2012-11-21 | 北京理工大学 | High-order extended Kalman filtering method |
Non-Patent Citations (3)
| Title |
|---|
| 卡尔曼滤波在跟踪运动目标上的应用;栗素娟等;《现代电子技术》;20071231(第13期);第110-112页 |
| 基于像素时域特征的红外弱小目标检测算法研究;沈强;《中国优秀硕士学位论文数据库信息科技辑》;20090715(第07期);全文 |
| 基于卡尔曼滤波的目标估计和预测方法研究;成光等;《计算机仿真》;20060131;第23卷(第1期);第8-10页 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105631063A (en) | 2016-06-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108362281B (en) | Long-baseline underwater submarine matching navigation method and system | |
| US10386476B2 (en) | Obstacle detection method and apparatus for vehicle-mounted radar system | |
| CN107084714B (en) | A kind of multi-robot Cooperation object localization method based on RoboCup3D | |
| CN101907459B (en) | Monocular video based real-time posture estimation and distance measurement method for three-dimensional rigid body object | |
| CN104236551B (en) | A kind of map creating method of snake-shaped robot based on laser range finder | |
| CN101408985B (en) | Method and apparatus for extracting circular luminous spot second-pixel center | |
| CN110197615B (en) | Method and device for generating map | |
| CN101183266A (en) | Method, apparatus, and medium for estimating pose of mobile robot using particle filter | |
| CN102609940A (en) | Method for processing errors generated by point cloud registration in process of surface reconstruction of measuring object by using ground laser scanning technique | |
| KR101888295B1 (en) | Method for estimating reliability of distance type witch is estimated corresponding to measurement distance of laser range finder and localization of mobile robot using the same | |
| CN113112510B (en) | Semantic segmentation forest fire detection method, controller and storage medium | |
| CN103759701B (en) | Based on the cell phone intelligent distance-finding method of Android platform | |
| CN111598942A (en) | Method and system for automatically positioning electric power facility instrument | |
| CN104331927A (en) | Method and system for simulating digital outcrop surface of reservoir | |
| CN109737968B (en) | Indoor fusion positioning method based on two-dimensional LiDAR and smart phone | |
| CN110763239A (en) | Filtering combined laser SLAM mapping method and device | |
| CN109870106A (en) | A kind of building volume measurement method based on unmanned plane picture | |
| CN105444766A (en) | Indoor navigation method based on deep learning | |
| CN113776519A (en) | AGV vehicle mapping and autonomous navigation obstacle avoidance method under lightless dynamic open environment | |
| CN105631063B (en) | A kind of moment forecasting procedure that crosses based on target signature | |
| CN109000638A (en) | A kind of small field of view star sensor measurement filtering wave by prolonging time method | |
| CN106408600B (en) | A method of for image registration in sun high-definition picture | |
| CN112711255B (en) | Mobile robot obstacle avoidance method, control equipment and storage medium | |
| CN111199191A (en) | Pointer instrument automatic reading method based on scale searching | |
| CN113160191B (en) | Environmental composition integrity judging method and device based on laser radar |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |