CN104049259A - Lidar three-dimensional imaging system based on virtual instrument - Google Patents

Abstract

A lidar three-dimensional imaging system based on a virtual instrument comprises a three-dimensional scene modeling module, a lidar testing system simulation environment modeling module, a full-waveform signal processing module and a three-dimensional reconstruction module. The three-dimensional scene modeling module comprises a three-dimensional model loading element used for setting background colors, visual angle control, illumination, a projection mode, a display model and other basic scene projects. The lidar testing system simulation environment modeling module comprises four modeling sub-modules which are a laser pulse model, an atmospheric transmission model, an object interaction model and a receiving unit model. The laser pulse model is used for simulating a laser source according to the wavelength, pulse width, energy and other characteristics of a laser. The atmospheric transmission model is used for simulating a tested atmospheric environment, so that a noise model is generated for acting on the laser. The object interaction model is used for simulating effects of the laser and a detected object, and the effects comprise mirror reflection, diffuse reflection, surface reflection and speckles. The receiving unit model is used for simulating detector noise and amplifier noise which are generated after sensing.

Description

Laser radar 3-D imaging system based on virtual instrument

One, technical field

The present invention is the laser radar 3-D imaging system based on virtual instrument, belongs to laser radar detection technical field.

Two, background technology

Laser radar detection and range measurement system (Light Detection and Ranging, be called for short Lidar) be a kind of active remote sensing technology, it has broken through traditional passive remote sensing imaging mechanism, be subject to that weather effect, illumination effect are little, automaticity high, there is unrivaled advantage.

Airborne imaging Lidar is by IMU aligning, DGPS acquisition device projection centre position, IMU/DGPS system refers to utilizes dress GPS receiver aboard and the GPS receiver being located on ground one or more base station synchronously and continuously to observe gps satellite signal, obtain the location parameter of aerial surveying camera by gps carrier phase measurement differential position, high precision Inertial Measurement Unit (the IMU that application and aerial surveying camera are closely connected, Inertial Measurement Uint) directly measure the attitude parameter of aerial surveying camera, pass through IMU, the associating post-processing technology of DGPS data obtains the photogrammetric measurement technology of every required photo high precision elements of exterior orientation of mapping, by Point Cloud Processing and filtering processing, three-dimensional reconstruction DEM.

In conjunction with multi-door ambits such as optics, radar, signal processing, computer visions, its range of application is across sea, land and sky for laser radar technique, is that recent domestic develops rapidly and the technical field of widespread use.Laser radar is applied to aerospace field the earliest, afterwards gradually for military affairs, fields such as laser guidance, battle reconnaissance, airplane anti-collision, land mine remote sensing.At present, laser radar technique is widely used in the various fields such as city digital Model Reconstruction, forest ecology monitoring, marine environment mapping, geology and geomorphology detection, space exploration.

The backscatter waveform of (Full-Waveform Lidar) the paired pulses transmitted waveform of all-wave laser radar and each pixel carries out data acquisition record, can decomposite flight time (TOF) and the strength information of each pixel on a time shaft.

The method of laser infrared radar imaging is mainly divided into three kinds:

(1) adopt single detector, only measure a pixel at every turn.This method is Detection Techniques the earliest, by the flight time of single measurement different pixels and strength information, recovers one by one distance and the reflectivity of each pixel.

(2) adopt planar array detector, each detectable multiple pixels.By utilizing emitted light is covered to whole target simultaneously, echoed signal contains the information of multiple pixels, and receiver can draw the distance of different pixels by demodulation, separation.This method is for large spot imaging, and more single of the pixel number of single measurement is measured obviously and improved.

(3) adopt APD (snow spring photodiode) detector array, the multiple pixels of standalone probe.Restraint on N the point of beating respectively in target by light beam being divided into N, receive by N unit of APD array, thus can the while information of a decomposing N point exactly.Scan all pixels of target by scanner, final imaging.The method has adopted highly sensitive APD sensor array, and two kinds of methods, can ensure high resolving power one-tenth figure above, can ensure again high hasty map.

Full wave shape echoed signal modeling and simulating is an important research content of Full wave shape Airborne Lidar survey technology with analyzing.The achievement of an actual measurement system often needs simulation modeling in earlier stage to carry out theoretical validation.About laser radar modeling and simulating aspect, there were a lot of successfully technological achievements abroad, and domestic to this rare systematic research achievement.

(1) the 3D Lidar system emulation of Lincoln laboratory based on OPENGL exploitation the hidden airborne 3D ladar test off the net of leafage, by interpretation of result, the result matching degree of simulation result and measured test is in 20%.

(2) Graham of Australia has carried out emulation based on Matlab to laser radar vegetation penetration test under carrying platform different motion track.

(3) Kim of the U.S., by setting up the Monte Carlo model of laser penetration trees, obtains Lidar full-wave simulation signal.This model, by Lidar system and natural scene, comprises atmosphere, and trees and landform finally obtain 3D model and the reflection characteristic of vegetation.

(4) German JUTZI, by the all-wave signal on the different gradients of emulation slope, utilizes and measures the gradient of determining plane with the correlativity of estimating ripple.

In existing technological achievement, not occurring can integrated control Lidar Simulation experiment scene, again can be for application platform provides hardware module interface, realize the Lidar 3-D imaging system of the expanded application of laser radar actual measurement experiment.Application of Virtual based on LabVIEW is a kind of brand-new exploitation implementation to laser infrared radar imaging field, by building of convenient system Visualization Platform to a great extent.LabVIEW is a graphical programming software based on G (Graphic) language.It adopts GRAPHICAL PROGRAMMING method, uses graphic symbolic expression program behavior, uses visualization technique to set up man-machine interface, more directly perceived, succinct.Abundant function library and subroutine that LabVIEW comprises as data acquisition, signal processing, I/O drive supervisor storehouse, provide multilingual development interface simultaneously, become the ideal tools of the application such as development and testing, measurement, data acquisition.

Three, summary of the invention

The object of the invention is the platform based on LabVIEW, develop the emulation test system of a set of complete laser radar three-dimensional imaging and the expansion of surveying 3-D imaging system based on the Lidar of virtual instrument is provided.

Laser radar detection based on virtual instrument and a range measurement system three-dimensional imaging analogue system, comprise three-dimensional scenic modeling, the modeling of Lidar pilot system simulated environment, the processing of all-wave signal, four functional modules of three-dimensional reconstruction; The feature of modules is described below:

(1) three-dimensional scenic MBM: read in three-dimension developing instrument 3DSMAX, the three-D elements of the forms such as the STL/WRL/ASE that AutoCAD produces, background colour is set, viewing angle control, illumination, projection mode, the basic scene project such as display model, to the landform that comprises of reading in, vegetation, the scope texture of aircraft, size, position, number is controlled, the wherein corresponding lasing light emitter in the position of aircraft position, dynamic change in test, finally present airborne Lidar imaging test dynamic scene, and export final digital elevation model DSM altitude figures, in order to follow-up test,

(2) Lidar pilot system simulated environment MBM: comprise laser pulse model, propagation in atmosphere model, target interaction models, four sub-module modelings of receiving element model; Laser pulse model is simulated lasing light emitter according to character such as sharp light wavelength, pulsewidth, energy; Propagation in atmosphere model is simulated the atmospheric environment of test, and generted noise model is to laser effect; Target interaction models, to laser and detection of a target effect simulation, comprises mirror-reflection, diffuse reflection, surface reflection, speckle; Noise of detector, amplifier noise that receiving element model produces after in order to analog sensed.By to the modeling of four model orders, the systematic parameters such as difference of elevation, temperature, flight angle are set, finally export the all-wave signal of each time period target respective pixel point;

(3) all-wave signal processing module: process the pixel degree of depth, the strength information information recovered for the echoed signal that each time period produces.This module mainly comprises signal enhancing, Wavelet Denoising Method, level and smooth, four parts of waveform fitting;

(4) three-dimensional reconstruction module: this module is tested final visual three-dimensional one-tenth figure is provided for airborne laser radar; The complete degree of depth and the strength information of target providing for all-wave signal processing module, generates Range Profile, intensity image; Be divided into non-ground point and ground point two category informations by segmentation and classification; Ground point information is set up grid, and ground point is carried out to triangle division, and aggregation strength information is finally rebuild generating digital elevation model DEM.

Further: in the corresponding different measuring cycle of aircraft position dynamic change in (1),, to contain pixel number in conjunction with hot spot and obtain pixel number and the echoed signal thereof that one-period records simultaneously laser light splitting according to APD number.

Further: (3) dynamically echoed signal search to each cycle, sampling unique point, specific aim processing, realizes the integrated function of signal enhancing, Wavelet Denoising Method, level and smooth, waveform fitting; In order to obtain higher measuring accuracy, waveform fitting module arranges matching template according to the actual waveform of laser pulse.

Further: based on the main development platform of LabVIEW, in conjunction with the mixed developing pattern of LabWindows/CVI, Matlab, C++, each function sub-modules (sub-VI) is called by father's module (father VI) according to inheritance, is finally integrated into the main gui interface of the superiors; Present the visual simulating test of complete laser radar three-dimensional imaging by carrying out total GUI.

Propose a kind of solution of the actual measurement of the Lidar based on virtual instrument imaging system, reserved TCP, serial communication interface, loaded measured data by collector, completed the function of Lidar eyeball cloud data three-dimensional imaging.

This platform utilizes the advantage of LabVIEW graphic language programming, uses dexterously three-dimensional value-added tax function control, builds adjustable laser radar experiment scene; To Lidar pilot system simulated environment, comprise that laser instrument, atmosphere, target are mutual, receiving element carries out comprehensive simulated, implement l-G simulation test and generate all-wave signal; Build the all-wave signal processing method of a cover system, all-wave signal is processed to resolve a picture dot point cloud information; Three-dimensional reconstruction function realizes module to cloud data classified filtering, final three-dimensional reconstruction DEM.

Beneficial effect of the present invention: virtual instrument framework: virtual instrument technique provides abundant high-performance modularized hardware interface, its Software tool LabVIEW provide pattern development platform, the simultaneously integrated abundant algorithms libraries such as data acquisition, signal processing, image processing that comprise, greatly shorten the construction cycle, there is very strong scalability simultaneously, realized software and hardware seamless integrated.Traditional upper imaging system of laser infrared radar imaging system is nearly all based on the platform such as Matlab, VC, realizing in experiments of measuring advantage not obvious.And it is very rare that virtual instrument is applied to laser infrared radar imaging field, and have Lidar visible, the system simultaneously with complete echo signal processing, imaging integrated treatment never occurs.

Three-dimensional visualization dynamic test scene construction: utilize LabVIEW tri-dimensional picture to load control, provide the three-dimensional model element form that professional cartography instrument is generated to load, by three-dimensional model deformation, Lidar l-G simulation test Visual Scene is built in the controls such as texture.Whole l-G simulation test is based on LabVIEW platform, can be in conjunction with multiple types of tools exploitations such as LabWindows/CVI, VC++, Matlab, in conjunction with three-dimension developing instruments such as 3DSMAX, AutoCAD, Pro/E, SolidWorks, VRMLPAD, by dissimilar 3D Unified Model and set up Lidar test scene, feature richness, extendability is strong, effect of visualization is good, complete and presented intuitively the test of all-wave laser radar three-dimensional imaging and functional verification is provided.Meanwhile, system provides the upper interface of actual measurement system, and to measured test, three-dimensional imaging provides the solution of expansion.

Four, brief description of the drawings

Fig. 1 is the structured flowchart of the imaging Lidar emulation platform based on virtual instrument;

Fig. 2 is that three-dimensional scenic module realizes three-dimensional modeling process flow diagram;

Fig. 3 is the process flow diagram that analogue system is set up;

Fig. 4 is the flow chart of data processing figure of all-wave signal processing module;

Fig. 5 is three-dimensional reconstruction module process flow diagram;

Fig. 6 is all-wave sampled signal waveform figure;

Fig. 7 is the imaging Lidar simulation software operational flowchart based on LabVIEW;

Fig. 8 is the main gui interface of the imaging Lidar simulation software based on LabVIEW;

Fig. 9 is three-dimensional scenic module modeling effect schematic diagram;

Figure 10 is the final three-dimensional reconstruction remodelling of Lidar l-G simulation test DEM;

Five, embodiment

Below in conjunction with accompanying drawing, implementer's case of the present invention is elaborated.

As shown in Figure 1, the present invention mainly comprises four functional modules: three-dimensional scenic modeling, the modeling of Lidar pilot system simulated environment, the processing of all-wave signal, three-dimensional reconstruction.Wherein three-dimensional scenic MBM can load the 3D model element of 3D mapping software output, and it is carried out to layout, adjusts, and makes it to produce required experiment scene.The establishment mode of 3D model element is to enrich flexibly.For example, the model element of landform can use the topography and geomorphology plug-in unit terrian plug-in unit in 3dsmax, generates the real figure landform of specifying longitude and latitude by connecting GIS; Vegetation model can utilize the tree plug-in unit structure of making of 3dsmax to specify vegetation group; Model aircraft can be at AutoCAD according to designing in dimensional requirement.

Shown in Fig. 2 is that three-dimensional scenic is built, and can load the 3D model file of 3D mapping software output, can obtain the implementation method of vertical array mould module by calling scene grid node.LabVIEW provides the information such as VRML, STL, tri-kinds of conventional 3-D geometric model loading methods of ASE, color array, texture coordinate, by setting up geometric model, add object, texture etc. is set calls nodal function and merge dissimilar 3D model buildings three-dimensional scenic.Meanwhile, LabVIEW provides abundant scene control method, comprises the basic setups such as background colour, light source, viewing angle control device, automatic projection mode, display mode, and the distortion control functionality such as scene convergent-divergent, translation, rotation.By merging these functionality controls in interface, foreground, can control flexibly three-dimensional scenic layout, finally present required emulation experiment scene, and export the digital elevation information of scene, use as target simulator experiment.In embodiment, setting aircraft flight height is 2km, and APD number is 8 array elements, and laser frequency is 10kHz, and atural object hot spot footprint is contained 8 pixel information.By the dynamic linear array scanning of for circulation control control aircraft, single sweep operation time corresponding real time 0.1ms.

Fig. 3 has provided Lidar pilot system simulated environment modeling embodiment, and this scheme is according to existing Laser beam energy distribution formula, the modeling of propagation in atmosphere formula.Lidar test simulation need for environment carries out modeling to laser pulse model, propagation in atmosphere model, target interaction models, receiving element model.The content of laser pulse modeling is the structure to Laser Time distribution, irradiance distribution, and input parameter comprises sharp light wavelength, half pulsewidth, the angle of divergence, laser energy, operating distance, frequency.The object of propagation in atmosphere model is the impact of simulation beam broadening, atmospheric turbulence, atmospheric attenuation, acts on laser pulse, and input parameter comprises spectral irradiance, turbulence intensity, visibility distance, luminance factor.The object of target interaction models is the impact that simulated laser pulse and the detection of a target are made mirror-reflection, diffuse reflection that the used time produces, surface reflection, speckle, and input parameter comprises specular reflection factor, the diffuse reflection factor, instantaneous ramp rate.Receiving element model is to want simulated laser effect target back echo signal to be subject to the noise effect of detector, amplifier through receiving trap, and the parameter of input comprises iris radius, instantaneous field of view, bandwidth, light transmission, responsiveness, focal length, dark face electric current, phaeodium electric current, excess noise fact or, detecting device internal gain, average background reflection, optical filter bandwidth, detector capacitors value, detecting device and amplifier total magnification.The correlation parameter of whole system model has difference of elevation, temperature, flight angle, target reflectivity, starts the each module of system simulation model, and order is carried out and produced Lidar Simulation echo waveform.In embodiment, (1) laser pulse time propagates irradiance distribution (2) propagation in atmosphere model comprises beam-broadening, atmospheric turbulence, energy attenuation: 1. beam-broadening rule: ${\mathrm{\ω}}_{\mathrm{spread}}=\mathrm{\ω}\·\sqrt{1+1.624\·{\left({\mathrm{\σ}}_{\mathrm{Rytov}}^2\right)}^{6/5}\·{\mathrm{\Λ}}_1},$ Wherein c _nfor textural constant, the λ of atmospheric turbulence are that optical maser wavelength, R are target object distance; 2. atmospheric turbulence rule: atmospheric scinillation probability density function is $P_{\mathrm{turb}}=\frac{1}{S\·\sqrt{{2\mathrm{\π\σ}}_{\ln I}^2}}\exp (-\frac{{(\ln \left(\frac{S}{S_{\mathrm{av}}}\right)+\frac{1}{2}{\mathrm{\σ}}_{\ln I}^2)}^2}{{2\mathrm{\σ}}_{\ln I}^2}),$ Wherein ${\mathrm{\σ}}_{\ln I}^2=\ln (1+{{\mathrm{\σ}}_I}^2)=\ln (1+{\mathrm{\σ}}_{t\arg \mathrm{et}}^2),{\mathrm{\σ}}_{t\arg \mathrm{et}}^2={\mathrm{\γ}}_{t\arg \mathrm{er}}\·{\mathrm{\σ}}_{\mathrm{sat},\mathrm{point}}^2={\mathrm{\γ}}_{t\arg \mathrm{et}}\·\frac{{1.3\mathrm{\σ}}_{\mathrm{point}}^2}{1+{\mathrm{\σ}}_{\mathrm{point}}^2}$ For the average rear effectively turbulent wave momentum of target, wherein r _eff=min (r _target, ω, υ _fov/ 2R) be effective light spot radius (r _targetfor the geometric radius of target, ω is the radius of laser at target place, υ _fov/ 2R is the visual radius of target of receiver), ρ _lfor (one dimension) effectively turbulent flow radius, determined by atmospheric turbulence constant C n, target range R, laser wavelength lambda and laser beam divergence φ; 3. energy attenuation rule: r is operating distance, σ _aerit is decay factor; (3) target interaction models comprises mirror-reflection, irreflexive impact, is expressed as: $\mathrm{\ρ}={\mathrm{\ρ}}_{\mathrm{spec}}+{\mathrm{\ρ}}_{\mathrm{diff}}=\frac{A}{{\cos }^6\left(\mathrm{\θ}\right)}{e}^{-\frac{{\tan }^2\left(\mathrm{\θ}\right)}{s^2}}+{B\cos }^m\left(\mathrm{\θ}\right),$ A and B are mirror-reflection and irreflexive association factor; (4) receiving element model is according to the noise equivalent power modeling of detector and amplifier: wherein e _cfor electron charge, B is bandwidth (Hz), I _dsfor surperficial dark current (A), I _dbfor dark current (A), I _bfor the electric current (A) that ground unrest produces, I _sfor marking current (A), M is detector internal gain, and F is excess noise factor, for the response of detector.

Fig. 4 is the all-wave signal processing module of laser infrared radar imaging system.The function implementation procedure of this module is: all-wave signal, by specifying sampling rate to obtain all-wave sampled data, carries out signal enhancing to this signal, as methods such as threshold value enhancing, low-pass filtering, pulse accumulation.It is 1GSPS that embodiment sets sampling rate, and single-pulse laser pulsewidth is 10ns, and frequency is 10kHz.Echo data for model generation is sampled according to 10 sampled point values of taking out of monopulse, finally obtains actual emulation all-wave sample waveform, as shown in Figure 7.Carry out noise pre-service through low-pass filtering, then get the noise sample of 8 APD echoed signals, carry out region accumulation, y ' (i)=Σ _na _ny _n(i), y _n(i) represent input sample signal, a _nrepresent weight, the signal to noise ratio (S/N ratio) of corresponding each group signal, n is sample number.

Ethernet host computer based on LabVIEW in embodiment is realized the input of actual measurement sampled data.Implementation method: call " opening TCP connects " function of labview, input server ip address, port address, by the transmission of three-way handshake command communication log-on data, completes the upper bit function of laser all-wave sampled data.The echo data position actual measurement laser all-wave sampled signal now obtaining.

All-wave sample waveform is carried out to noise pre-service, smoothing processing.Embodiment adopts wavelet de-noising disposal route.Wavelet de-noising has good time-frequency characteristic, selects base flexible, can process for ground the detailed information such as each edge, spike of non-stationary signal.In embodiment, call LabVIEW small echo kit, select " Advanced Signal Processing Toolkit → Wavelet Analysis → Feature Extraction → Wavelet Denoise " to call Wavelet Denoising Method function.Can select suitable de-noising level, threshold value selective rule and wavelet basis, input signal is processed.De-noising level is generally selected 4-5 level, threshold value is selected common several soft-thresholds, as without evaluation of risk threshold value (rigrsure) partially, to inspire three kinds of threshold value (heursure) and minimax threshold values (minimaxi), wavelet basis be mainly tetra-kinds of sym, db, bior, coif.According to actual effect, adopt 5 grades, rigrsure, tri-parameters of sym8 can reach comparatively desirable effect.Smoothing algorithm is in order to signal deburring, can select the method for moving average, 53 smoothing algorithms, vondrak algorithm in embodiment.Wherein vondrak algorithm smooth effect is best, effectively filtering burr, and the while can ensure again the integrality of signal well.Vondrak smoothing method is that Czech astronomer Vondrak proposes, and is applicable to uniformly-spaced be suitable for again the smoothing algorithm of unequal interval., its basic level and smooth criterion is: Q=F+ λ ²s=min, wherein $F=\mathrm{\Σ}{p}_i{[x^{\′}\left(t_i\right)-x\left(t_i\right)]}^2,s=\underset{i=1}{\overset{N-3}{\mathrm{\Σ}}}{\left[{\mathrm{\Δ}}^3{x}^{\′}\left(t_i\right)\right]}^2,$ X (t _i) for treating smooth value.

Peak value, flex point detect for the number of definite waveform decomposition and initialized location and the peak value of nonlinear fitting, need before this to delete the wrong point of erroneous judgement, and calculated threshold are as matching constant.Implementation method: for peak value, search first order derivative=0; For rising edge flex point: first order derivative >0, second derivative=0, three order derivative <0; For negative edge flex point: first order derivative <0, second derivative=0, three order derivative >0.Nonlinear fitting is the core algorithm that waveform decomposes, and method is many, for example Levenberg-Marquardt nonlinear fitting, and LabVIEW provides this functionality controls directly to call.What existing approximating method used is Gauss curve fitting, there is certain error for laser pulse matching meeting.Actual Laser Time is distributed as: wherein t _1/2represent half-wavelength.Therefore, use as LM matching template, can obviously improve the goodness of fit, reduce error of fitting.Decompose the TOF (flight time), the strength information that obtain respective pixel by waveform.In embodiment, call " nonlinear fitting of belt restraining " function in LabVIEW, selected " LM boundary " pattern, sets matching template according to above-mentioned form, because adopting the pulse of 10ns pulse width laser, therefore set initialization condition a (3i+1)=peak-2a (3i+2), a (3i)=peak amplitude-4e ^-2, utilize LM iterative algorithm, output cycle index, best fit parameters.According to formula digital simulation goodness.Under suitable starting condition input, the goodness of fit can reach more than 0.999, can be used as final waveform decomposition result.To each wavelet of decomposing, its time delay is the elevation information of corresponding pixel points, and its amplitude is the strength information of corresponding pixel points.

Fig. 5 is the three-dimensional reconstruction module of Lidar imaging system.The implementation of this module is: utilize the elevation information of cloud data decomposition and the two-dimensional image of strength information generation target, as Range Profile, intensity image; Process is cut apart, is classified pixel, and filtering obtains non-ground point and ground point.Set up three-dimensional coordinate, generating mesh, to coordinate point cloud triangulation, merges degree of depth picture and intensity image, finally obtains digital elevation model.

Fig. 6 is the operating process of the laser radar 3-D imaging system based on LabVIEW, and its corresponding User Interface (GUI) as shown in Figure 7.

The first step: start after software platform, three-dimensional scenic basic parameter is carried out to initial configuration, load the three-dimensional model of landform, vegetation, aircraft, the situation elements generating is carried out to the controls such as size, position, texture, comprehensive layout, finally generates and presents Lidar three-dimensional scenic.

Second step: laser model parameter, Atmospheric models parameter, detector model parameter are set, and global parameter, for producing scenario triggered test, obtains and present all-wave signal.

The 3rd step: trigger the processing of all-wave signal, real-time resolving goes out the pixel depth information of each time cycle, and refreshes demonstration, can trigger data file.

The 4th step: trigger point cloud three-dimensional reconstruction, generate DEM and circle of equal altitudes, it can be set and present the demonstration of window shape formula, and can be to picture archiving.

The test effect that GUI platform based on Fig. 7 produces is as Fig. 8, shown in Fig. 9.Fig. 8 be three-dimensional scenic module construction in Fig. 7 Lidar l-G simulation test scene present window, Fig. 9 is three-dimensional DEM figure and the circle of equal altitudes that three-dimensional reconstruction module generates.

The above is embodiments of the invention, does not limit the present invention, all any amendments of making within the scope of thought of the present invention and principle, is equal to and replaces or invalid insertion etc., all should be included within protection scope of the present invention.

Claims (7)

1. the laser radar 3-D imaging system based on virtual instrument, is characterized in that comprising three-dimensional scenic modeling, the modeling of Lidar pilot system simulated environment, the processing of all-wave signal, four functional modules of three-dimensional reconstruction;

(1) three-dimensional scenic MBM: be written into three-dimensional model element, the basic scene project such as background colour, viewing angle control, illumination, projection mode, display model is set, the scope texture, size, position, the number that comprise landform, vegetation, aircraft that read in are controlled, the wherein corresponding lasing light emitter in the position of aircraft position, dynamic change in test, finally present airborne Lidar imaging test dynamic scene, and export final DSM surface altitude figures, in order to follow-up test;

(2) Lidar pilot system simulated environment MBM: comprise laser pulse model, propagation in atmosphere model, target interaction models, four sub-module modelings of receiving element model; Laser pulse model is simulated lasing light emitter according to character such as sharp light wavelength, pulsewidth, energy; Propagation in atmosphere model is simulated the atmospheric environment of test, and generted noise model is to laser effect; Target interaction models, to laser and detection of a target effect simulation, comprises mirror-reflection, diffuse reflection, surface reflection, speckle; Noise of detector, amplifier noise that receiving element model produces after in order to analog sensed.By to the modeling of four model orders, the systematic parameters such as difference of elevation, temperature, flight angle are set, finally export the all-wave signal of each time period target respective pixel point;

(3) all-wave signal processing module: process for the echoed signal that each time period produces, recover the pixel degree of depth, strength information.This module mainly comprises signal enhancing, denoising, level and smooth, four parts of decomposition;

(4) three-dimensional reconstruction module: this module is tested final visual three-dimensional one-tenth figure is provided for airborne laser radar; The complete degree of depth and the strength information of target providing for all-wave signal processing module, generates Range Profile, intensity image; Be divided into non-ground point and ground point two category informations by segmentation and classification; Ground point information is set up grid, and ground point is carried out to triangle division, and aggregation strength information is finally rebuild generating digital elevation model (DEM).

2. the laser radar 3-D imaging system based on virtual instrument according to claim 1, it is characterized in that the corresponding different measuring cycle of aircraft position dynamic change in (1),, cover pixel number in conjunction with hot spot and obtain pixel number and the echoed signal thereof that one-period records simultaneously laser light splitting according to APD number.

3. the laser radar 3-D imaging system based on virtual instrument according to claim 1, it is characterized in that in (3) dynamically the echoed signal search to each cycle, sampling unique point, specific aim processing, realizes the integrated function of signal enhancing, denoising, level and smooth, waveform fitting; In order to obtain higher measuring accuracy, waveform fitting module arranges matching template according to the actual waveform of laser pulse.

4. according to the laser radar detection based on virtual instrument described in claim 1 or 3 and range measurement system three-dimensional imaging analogue system, it is characterized in that reserved TCP, serial communication interface, load measured data by collector, complete the function of Lidar eyeball cloud data three-dimensional imaging.

5. according to the laser radar detection based on virtual instrument described in claim 1 or 3 and range measurement system three-dimensional imaging analogue system, the all-wave signal processing module implementation procedure that it is characterized in that imaging system is: all-wave signal is by specifying sampling rate to obtain all-wave sampled data, this signal is carried out to signal enhancing, threshold value enhancing, bandpass filtering, pulse accumulation method, noise reduction module is processed the detailed information such as each edge, spike for ground; Smoothing algorithm, to the level and smooth deburring of denoised signal, ensures the integrality of signal characteristic simultaneously.Peak value, flex point detect determines the number of waveform decomposition and initialized location and the peak value of nonlinear fitting, based on Levenberg-Marquardt Nonlinear Quasi hop algorithm, using practical laser pulse waveform as LM matching template, obviously improve the goodness of fit, reduce error of fitting.Decompose the depth information, the strength information that obtain respective pixel by waveform.

6. according to the laser radar 3-D imaging system based on virtual instrument described in claim 1 or 3, it is characterized in that the main development platform based on LabVIEW, utilize LabWindows/CVI, the Matlab of ActiveX technology, the mixed developing pattern of C++, build user interactions function of tonic chord interface.Can present the visual simulating test of complete laser radar three-dimensional imaging by carrying out main GUI, final three-dimensional one-tenth figure.Operating process is:

The first step: start after software platform, three-dimensional scenic basic parameter is carried out to initial configuration, load the three-dimensional model of landform, vegetation, aircraft, the situation elements generating is carried out to the controls such as size, position, texture, comprehensive layout, finally generates and presents Lidar three-dimensional scenic;

Second step: laser model parameter, Atmospheric models parameter, detector model parameter are set, and global parameter, for producing scenario triggered test, obtains and present all-wave signal;

The 3rd step: trigger the processing of all-wave signal, real-time resolving goes out the pixel depth information of each time cycle, and refreshes demonstration, can trigger data file;

The 4th step: trigger point cloud three-dimensional reconstruction, generate DEM and circle of equal altitudes, it is set and presents window display format, to picture archiving.

7. laser radar detection and the range measurement system three-dimensional imaging analogue system based on virtual instrument according to claim 5, the all-wave signal processing module implementation procedure that it is characterized in that imaging system is: noise reduction has good time-frequency characteristic, select base flexible, process the detailed information such as each edge, spike of non-stationary signal for ground.Peak value, flex point detect for the number of definite waveform decomposition and initialized location and the peak value of nonlinear fitting, need before this to delete the wrong point of erroneous judgement, and calculated threshold are as matching constant; Nonlinear fitting is that core algorithm that waveform decomposes adopts and comprises and provide this functionality controls directly to call by LabVIEW by Levenberg-Marquardt nonlinear fitting; Use Gauss curve fitting, there is certain error for laser pulse matching meeting; Actual Laser Time is distributed as: wherein t _1/2represent half-wave; Therefore, use as LM matching template, obviously improve the goodness of fit, reduce error of fitting.Decompose the TOF (flight time), the strength information that obtain respective pixel by waveform.