CN101901501B - Method for generating laser color cloud picture - Google Patents

Method for generating laser color cloud picture Download PDF

Info

Publication number
CN101901501B
CN101901501B CN2010102372655A CN201010237265A CN101901501B CN 101901501 B CN101901501 B CN 101901501B CN 2010102372655 A CN2010102372655 A CN 2010102372655A CN 201010237265 A CN201010237265 A CN 201010237265A CN 101901501 B CN101901501 B CN 101901501B
Authority
CN
China
Prior art keywords
line
laser
laser scanner
digital camera
scan digital
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
Application number
CN2010102372655A
Other languages
Chinese (zh)
Other versions
CN101901501A (en
Inventor
钟若飞
宫辉力
王留召
刘先林
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Capital Normal University
Original Assignee
Capital Normal University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Capital Normal University filed Critical Capital Normal University
Priority to CN2010102372655A priority Critical patent/CN101901501B/en
Publication of CN101901501A publication Critical patent/CN101901501A/en
Application granted granted Critical
Publication of CN101901501B publication Critical patent/CN101901501B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Abstract

The invention provides a method for generating a laser color cloud picture, which is to set a plurality of target points in a scene to be detected. The method comprises the following steps: S10, rigidly binding a line camera and a laser scanner in parallel to ensure that two scanning surfaces of the line camera and the laser scanner are parallel, and the scanning surfaces are vertical to a line between the front node of the camera and the center of the laser scanner; S20, leading the line camera and the laser scanner to scan synchronously; S30, establishing a corresponding relationship between a scanning angle of the corresponding target point and pixel sequence number on a line image, i.e. a color searching list; and S40, establishing a relationship between each laser point in laser dot cloud and RGB information thereof according to the color searching list and the scanning time difference between the laser scanner and the line camera. The method ensures that fusion of the dot cloud and the image is changed from three-dimensional fusion in the prior art to polar coordinate one-dimensional fusion so as to reduce operation capacity, greatly improve the modeling accuracy, and achieve multiple functions. The method can greatly reduce time cost and economic cost of urban three-dimensional modeling.

Description

A kind of method that generates laser color cloud picture
Technical field
The present invention relates to a kind of method that generates laser color cloud picture, generate laser color cloud picture through laser scanning and line-scan digital camera more precisely, to be used for color three dimension image reconstruction etc.The present invention can be applicable to space flight/airborne remote sensing and photogrammetric and airborne/fields such as vehicle-mounted traverse measurement/spatial Information Service.
Background technology
Present objective reconstruction technique relates generally to the profile of objective and confirms to fill two aspects with texture.Its major technique approach has two: the one, and through the picture image data, utilize complicated large-scale calculations to carry out three-dimensional modeling; The 2nd, obtain outline data through means such as high-precision laser scannings earlier, obtain the color information of body surface again through camera head; Through the mode of data fusion, color information and outline data are mated at last, form texture mapping.
Mostly the method that adopts in the three-dimensional city modeling is article one approach.Be the basis with the close shot photo, utilize 3 d modeling software, use the mode of rebuilding Building Modeling.Aspect the buildings elevation obtains, generally, perhaps utilize measurement means point-to-point measurements such as total powerstation according to the artificial estimation of buildings floor height.This modeling method not only efficient is low; And precision is not high, is difficult to obtain buildings real space information, and these models can be used to three-dimensional range to be showed; But can't carry out spatial operation, measurement based on these models, this has become the bottleneck that city space information is used.
And the second approach; Because profile information and camera head work alone respectively, wherein the process of data fusion relates to edge of image extraction, profile identification, a large amount of image processing process such as object coupling; And receive the influence of algorithm, fusion precision also is difficult to control.
For example; Application number CN02116507.6; The Chinese invention patent of by name " Computerized 3 D visual color scan system and scan mode thereof " promptly discloses a kind of Computerized 3 D visual stereo colour scanning system, and this system adopts normal optical to scan the identification with the three-dimensional model profile of target object, utilizes the external form profile configuration of target object; Carry out the reconstruction of three-dimensional model; Every pictures that camera head is photographed superposes based on system coordinate system, generates the basic external form of object, promptly generates three-dimensional model.In profile identification, obtain the color information of body surface through camera head, promptly obtain the three-D grain pinup picture of scanned object.Both are synthetic, have promptly formed preliminary three-dimensional colour model.Utilize technique of laser imaging, utilize the principle of range of triangle, obtain the accurate coordinate value of body surface measurement point, generate a large amount of three-dimensional geometry data thus.The automatic superposition of data of normal optical scanning and laser scanning generation is compound, carry out the automatic superposition of computing machine in the data that diverse location scans out, thereby obtain accurately perfect three-dimensional data.The method has reflected the typical thinking of prior art when the object color information of spatial positional information that laser scanning is obtained and video camera shooting merges---will scan or take independent separately the completion; And respectively with the data that obtain (some cloud or 3 d structure model and the photograph image rebuild according to this) as independent integral body, carry out the identification and the coupling of object at last and fill texture.For certain object, it is all inequality that video camera and laser scanner obtain angle and the ratio of data etc. separately, and its coupling relates to a large amount of view data such as demarcation, coupling, translation, convergent-divergent, rotation and handles.
For another example; Application number is CN200910115564.9; The Chinese invention patent that the applying date is on 06 19th, 2009, by name " three-dimensional rebuilding method that merges based on laser and camera data " promptly discloses a kind of three-dimensional rebuilding method based on laser and camera data fusion.This method for reconstructing is: adopt that C Plus Plus writes based on the automatic acquisition platform of Principles of Radar laser data; Obtain rotation matrix R and translation vector T between camera coordinate system and the laser instrument coordinate system; Utilize the optical flow field region merging algorithm that moving target is extracted and split, and then the some cloud that moving target is corresponding splits; For the mapping of the three-D grain of image provides the assurance on accurate location of some cloud and image texture, thereby realization is based on the three-dimensional reconstruction system of data fusion.
The method improves to some extent than a last method; But the outside that still must carry out laser and CCD is demarcated; And relate to the available point cloud that extracts and cut apart target in the complex environment; Because the CCD that adopts is the picked-up of carrying out the planar image, it still belongs to laser data and the discrete collection of CCD bidimensional image data in essence, carries out data fusion method again.
To sum up, existing three-dimensional modeling method is difficult to satisfy the demand of large-scale city modeling owing to the restriction of data obtaining means and data processing method.
Summary of the invention
The objective of the invention is to, providing can a kind of method of carrying out three-dimensional reconstruction rapidly and accurately, and it is not high slow with modeling speed to solve in the prior art in the three-dimensional reconstruction process precision, the problem that data volume is big.
More specifically; The invention is intended to provide a kind of three-dimensional rebuilding method that combines based on laser scanning and CCD imaging; Corresponding fast and accurately can between the three dimensional local information of image and color information, setting up; Thereby reduce texture calculation of Matching amount, or even can realize that texture matees automatically.And, can avoid the demarcation of the external parameter of CCD camera to a certain extent.Thereby, make the color three dimension modeling can be in high precision, fast, accomplish under the little prerequisite of data volume and calculated amount.This will make the present invention be particularly suitable for the city three-dimensional modeling; Consider that particularly method of the present invention is based on " very three-dimensional " city modeling of actual measurement laser absolute fix information; Compared incomparable advantage with three-dimensional reconstruction based on two-dimensional image information; Not only embodied in three-dimensional modeling thought and technical progress, also will inevitably bring huge economic interests and the construction and the urban digital process of GIS-Geographic Information System played great impetus.
For this reason, core thinking of the present invention be select the line sweep CCD camera similar with laser scanner but not in the prior art generally the plane CCD camera of use get texture information.Utilize the high-precision characteristics of line-scan digital camera simultaneously; Make the some cloud of laser scanning can be endowed the RGB value of color that line-scan digital camera is gathered; Mapping through a cloud and RGB color value; Realize the geometric properties data of laser point cloud and, generate laser color cloud picture, and then realize the Fast Reconstruction of sense of reality building model from the automatic fusion of the texture information of CCD.And utilizing laser scanner and linear array CCD camera all is characteristics of line sweep, guarantees that through physical means in earlier stage the sweep trace of laser is parallel with the sweep trace of line-scan digital camera, sets up the simple mapping relations between some cloud and the pixel.
Particularly; The method of generation laser color cloud picture of the present invention may further comprise the steps: S10; Line-scan digital camera and laser scanner are carried out the parallel binding of rigidity, so that the scanning plane of the two is parallel, and these scanning planes are perpendicular to the line at camera front nodal point and laser scanner center; S20 makes line-scan digital camera and laser scanner carry out synchronous scanning; S30 sets up laser scanning angle and its corresponding relation---color lookup table between the pixel sequence number on the linear array image of impact point of the same name; S40 sets up each laser point and its RGB information corresponding relationship in the laser point cloud according to difference sweep time of color lookup table and laser scanner and line-scan digital camera.
Wherein, among the step S1, it is may further comprise the steps that line-scan digital camera and laser scanner are carried out the parallel binding of rigidity: S11, with the center-aligned of the front nodal point and the laser scanner of line-scan digital camera; S12, the angle of adjustment laser scanner and camera makes the two " imaging line " parallel; S13, the subtended angle of adjustment laser scanner and camera makes the constant distance between the two " imaging line "; S14 carries out the rigidity binding with line-scan digital camera that regulates and laser scanner.
Preferably; Among the said step S10, be with line-scan digital camera through having the double-deck rebound of three setting nuts, be installed in parallel in the laser scanner summit directly over; And between said setting nut, be provided with clearance gauge, carry out the parallel adjusting of the two through regulating clearance gauge.
According to one embodiment of the invention, the method for the angle of adjustment laser scanner and camera is: make laser scanner and camera at the uniform velocity mobile synchronously with speed V, a plurality of impact points that are arranged at the ring-type metope are carried out to picture; Calculate laser scanning line and camera sweep trace respectively through the mistiming Δ t of the some point1 in the impact point with some point2 L, Δ t C, the clearance gauge distance that then need adjust is: ds * Δ R i, wherein, Δ R i=alpha-beta, α=Δ t L* V/d, β=Δ t C* V/d, d are a point1 and the air line distance of putting point2.
According to one embodiment of the invention, the method for the subtended angle of adjustment laser scanner and camera is: make laser scanner and camera at the uniform velocity mobile synchronously with speed V, a plurality of impact points that are arranged at metope are carried out to picture; The perpendicular bisector of obtaining the monumented point line arrives the vertical apart from L of projection centre, obtains the elevation Z of the corresponding laser scanner of each monumented point then CiWith camera elevation Z Ji, said elevation is the two distance apart from starting point when monumented point is sampled, two elevations subtract each other and subtract the adjustment component v that camera front nodal point and the distance h (is example with 170mm) on laser scanner summit obtain corresponding each monumented point again i,
Figure GSB00000665329800041
needs the thickness=2 * ds * μ of adjustment; Wherein, ds is a clearance gauge thickness;
Figure GSB00000665329800042
N is a monumented point number in the image.
Wherein, it is to be fixed together with the scanning plane and the line-scan digital camera of mechanical mode with laser scanner that said rigidity is bound, and makes it can not relative motion; Perhaps laser scanner and line-scan digital camera are carried out synchronous follow-up control with the mode of mechanical/electrical.
According to one embodiment of the invention, said step S20 line-scan digital camera and laser scanner carry out synchronous scanning to be realized through the two being fixed on the controlled hoistable platform of a precision.
Preferably, among the said step S30, the foundation of color lookup table is according to as a token of identical point coordinates and the coupling of image information on the laser point cloud; At first; In a cloud, identify monumented point; Find the angle of corresponding point in the corresponding polar coordinates file afterwards according to gps time information subsidiary in the coordinate of monumented point, in image, measure the row number of this monumented point simultaneously, i.e. which pixel in its corresponding line-scan digital camera; Thereby set up the mapping relations between the corresponding polar angle of pixel of camera with laser scanner.
Preferably, laser scanning angle and its corresponding relation between the pixel sequence number on the linear array image are non-linear in the said color lookup table.
Preferably; Among the said step S40; Utilize gps data; Obtain camera and the two speed V (t) on direction of motion of laser scanner that rigidity is bound in real time; Try to achieve camera and the required mistiming t2-t1 of laser scanner displacement d through speed V (t) according to formula
Figure GSB00000665329800043
, the correspondence of setting up laser scanning line and camera thus pushes away the mapping of pipeline purging, and wherein d is the distance between camera front nodal point and the laser scanner center.
Beneficial effect of the present invention is, through the foundation of CCD line-scan digital camera with parallel binding of the rigidity of laser scanner and color lookup table, makes the fusion of invocation point cloud and image become the polar coordinate one-dimensional fusion by three-dimensional fusion the of the prior art.Operand is reduced to the degree of arithmetic by the complex figure processing, thereby has improved arithmetic speed greatly; And method of the present invention also makes the degree of accuracy of modeling that qualitative leap has compared with prior art been arranged owing to be that each laser point in the cloud is directly corresponding with the CCD pixel.Be based on the saving of three-dimensional modeling precision, modeling speed and data operation cost aspect, make method of the present invention can be applied to large-scale city three-dimensional modeling, bring great convenience to the foundation of urban digital construction and GIS-Geographic Information System.Thereby application of the present invention will bring remarkable social benefit and economic benefit.
Description of drawings
Fig. 1 is the schematic flow sheet of laser color cloud picture generation method of the present invention;
Fig. 2 is the process flow diagram of color lookup table one embodiment of the present invention;
Fig. 3 is the parallel binding method schematic flow sheet with the rigidity of laser scanner of line-scan digital camera of the present invention;
Fig. 4 is the synoptic diagram that angle is regulated in the parallel binding method of the present invention;
Fig. 5 is the synoptic diagram that subtended angle is regulated in the parallel binding method of the present invention.
Embodiment
Referring to Fig. 1, it has summarized the key step of laser color cloud picture generation method of the present invention on the whole.
At first; In step 10; Line-scan digital camera and laser scanner are carried out the parallel binding of rigidity; But its objective is corresponding relation relative fixed and analytical Calculation between the some cloud of the imaging pixel that makes line-scan digital camera and laser scanner, constant when the mapping relations that for example make the two are linear or non-linear with ideal, thus the data processing in later stage simply but can obtain very accurate result.
So-called " parallel binding " is meant between capture " thing " line (also claiming the camera sweep trace) and the determined plane of its front nodal point (for ease of narrating, this plane being called " camera scanning plane " among the present invention) of scanning plane and line-scan digital camera of laser scanner parallel.Preferably make these two parallel planes can be again perpendicular to the line between these 2 of the front nodal points of the laser center of laser scanner and line-scan digital camera.As for how realizing " parallel binding ", will combine Fig. 3-4 to specify at the back literary composition.
So-called rigidity is bound; In case be meant after the binding; Relative position relation between the scanning plane of laser scanner and the scanning plane of line-scan digital camera is constant; Its simplest embodiment is exactly to be fixed together with the scanning plane and the line-scan digital camera of mechanical mode with laser scanner, makes it can not relative motion.Certainly, with the mode of mechanical/electrical the scanning plane of laser scanner and line-scan digital camera are carried out synchronous follow-up control and also be fine, just cost is higher in realization.In case the benefit that the mechanical fixation formula is bound be bind after, if accidents such as no mechanical fault generally can not squint, probability of survival is higher, in case but break down then maybe not can by timely discovery.The advantage of controlled synchronous follow-up then is can find fault at any time because real-time verification is arranged, in time corrects, but slightly inferior aspect cost and reliability.Those skilled in the art can consider to select binding mode voluntarily according to many-sides such as data acquisition reliability and precision and costs, or conditions permit, and the mechanical fixation formula binding that has synchronization check will be the most reliable mode.
In step 20, make line-scan digital camera and laser scanner carry out synchronous scanning.After rigidity is bound, in the process of scanning modeling, moving of line-scan digital camera and laser scanner will be synchronous, i.e. " thing " line of sweep trace and capture is two synchronization-moving straight lines, so constant or its speed time-varying function v in the two translational speed L(t) and v CWhen being known (t), can set up the corresponding relation between the two very easily.Can realizing through the external trigger that laser scanner sends synchronously of camera and laser scanner; Because the data acquisition of laser scanner relates to a large amount of mechanical motions such as rotation of reflecting prism; Line-scan digital camera is then owing to mainly be that the running reaction of electronic component is fast and sensitive, thus by laser scanner send external trigger make camera with it synchronous operation implement more or less freely and reliable.
Also extremely important in the process of the parallel binding of synchronous scanning before system's operation.Because camera and laser scanner all are line sweeps, have only through moving to become two dimensional image.The synchronous of line-scan digital camera and laser scanner also is to mate through the same place between target image and the some cloud to find the solution the basis that the parallel binding with the realization of laser scanner alignment error angle of camera is regulated.For the ease of calculating; The precise guide rail of special use capable of using strict control camera and laser scanner (can certainly from top to bottom) from lower to upper at the uniform velocity push away and sweep; On the sensor image that obtains, seek same place respectively; According to the coordinate difference of same place on two sensor images, find the solution the alignment error angle of camera and laser, as the foundation of adjusting with accurate clearance gauge.
So-called " same place " is meant same imaging point corresponding respectively in laser scanner and two sensors of line-scan digital camera in the true environment.Can utilize significant nature or artificial environment mark to construct the same place that is easy to discern, " cross " mark that for example color, brightness and surrounding environment contrast are big.
In step 30, set up laser scanning angle and its corresponding relation---color lookup table between the pixel sequence number on the linear array image of impact point of the same name.
After the parallel rigidity binding of camera with the laser scanner process; For as the intersection between laser scanning face and camera scanning plane and the scanned object (the for example metope of buildings); Laser scanning line and camera sweep trace maybe be at different time through same positions; But by scanning " thing " line, the corresponding relation between two sweep traces is constant for same.This is because laser scanner is to rigidly fix to be in the same place with camera; Do not change and change with the position of carrier and attitude; So, can be based upon angle scanning direction and pairing camera in the laser scanning face corresponding relation between being listed as number, this mapping relations are referred to as color lookup table.When number is identical between the pixel of the angular resolution of laser scanner and line-scan digital camera, can set up the one-to-one relationship between the two; Number can be set up approximate corresponding relation through modes such as interpolation not simultaneously.Being noted that " table " this appellation, just for the correspondence between two groups of data of expression of image, and is not that representation or data storage form to said mapping relations limits.Certainly, can express this mapping relations with correspondence table, also available functions is expressed its mapping relations.On the data structure of storage, also can take any data structure of being convenient to read and write and storage mode according to the demand of program composition or Application of Interface.
If possible, can make in its laser point cloud the polar pixel count of counting of getting on each sweep trace suitable through the modes such as prism revolution of adjustment laser scanner with line-scan digital camera CCD, for example consistent or be integral multiple than relation.Inconsistent situation for most of the time is then calculated through the method for interpolation.
The establishment step of look-up table is according to as a token of identical point coordinates and the coupling of image information on the laser point cloud equally.At first; In a cloud, identify monumented point; Find the angle of corresponding point in the corresponding polar coordinates file afterwards according to gps time information subsidiary in the coordinate of monumented point, in image, measure the row number of this monumented point simultaneously, i.e. which pixel in the line array CCD (non-working direction).For example; If the pixel count of each line of line array CCD is 2048, color lookup table can be defined as sets up the mapping relations between the corresponding polar angle with laser of each pixel in the camera 1~2048, and the pixel of claiming camera among the present invention just is mapped as to the angle of laser shines upon; It is spaced apart 1 pixel; This look-up table is called positive look-up table, and with the inverse mapping that is mapped as to camera pixel of the angle of laser point, this look-up table is called anti-look-up table.
Can use all to help setting up the sign of corresponding relation between camera and the laser for measuring " look-up table " used target; Than the entanglement in mutually perpendicular traffic reflector plate, tennis court, the railing in flower nursery; Require the line-scan digital camera field angle the monumented point point symmetry cover within the visual field, if can not once cover, can move forward and backward the parking stall; Cover several times, set up the color lookup table in the full field range.Correct corresponding with accurately in order to set up, monumented point requires evenly to distribute to cover in the 0-2047 pixel scope, and at least 50 to 100 more than the point, and more monumented point will be favourable.
Utilize look-up table, can set up the relation (non-working direction) between laser point and the line-scan digital camera row direction, the relation between laser point and the row number (working direction) is set up through the mode of time correlation, obtains color cloud picture thus.After obtaining color cloud picture, if desired, can introduce checking procedure,,, can return at any time and rebulid look-up table if find in a cloud merges, to have error with the color cloud picture and the real image contrast of information fusion gained.
Here the accuracy of look-up table depends primarily on two aspects: the one, and what early stage, the camera collimation was adjusted is accurate; The 2nd, when look-up table is set up manual work measure same place between laser and the camera accurately whether.If out of true is with the effect that directly causes laser point cloud and camera color to merge.
Accompanying drawing 2 is diagrams of the color lookup table set up according to the method for the invention.Wherein horizontal ordinate is the pixel number of line-scan digital camera, and ordinate is the polar angle of laser point on the corresponding point cloud.In this diagram, laser scanner per second 100K, 50 circles; 100K/50=2000 point of every circle, the angle intervals between every is the 360/2000=0.18 degree, camera; 2048 points of every line, the CCD cell size is 14um, focal length is 14mm; The field angle of camera is that ((1024 * 14um)/14mm)=91.3587295 degree, the angle intervals between every in the camera is α/2048=0.044608746 degree to α=2 * arctg.Thus, between per two laser point, can in insert nearly 4 pixels, set up the mapping relations of laser point between camera, laser with the zone of viewing field of camera coupling in, whenever spend at a distance from 0.04, set up a pair of mapping.
In theory, look-up table is more carefully good more, because line-scan digital camera does not pass through accurate calibration, its elements of interior orientation, radial distortion etc. all are unknown, and one of purpose of setting up look-up table is exactly to avoid calibration in the complicated camera.Therefore, as can be seen from Figure 2, look-up table is pairing to be a nonlinear curve.
At last, in step 40, according to color lookup table and sweep time difference set up each laser point and its RGB information corresponding relationship in the laser point cloud.
Set up color lookup table; Solved each line laser scanner data and line-scan digital camera coupling in the direction of scanning; In actual three-dimensional imaging process, also to consider to push away the coupling of sweeping between last each the bar sweep trace of direction (pushing away and sweeping direction is the direction perpendicular to sweep trace generally speaking).Concerning the cloud atlas image, if the definition direction of scanning is the X axle, then pushes away and sweep direction and be generally the Y direction vertical with the X axle, in the three-dimensional modeling of city, said push away sweep direction can be consistent with the garage direction.And in calibration process in early stage such as parallel binding, push away and sweep direction and be chosen as vertical direction, and with target 360 degree around setting.
Corresponding which the bar camera of which bar laser scanning line of direction is swept by garage or push away pushes away pipeline purging and can lean on the time association.After calibration, the center of laser scanner is between the camera front nodal point being known definite value apart from d.If the known again speed V (t) that sweeps and the acquisition time of laser scanner and camera scan-data of pushing away just can the corresponding relation of same impact point between the data of the two collection.
Describe for simplifying; If line-scan digital camera scanning is t1 through the time of some P1, laser scanning is t2 through the time of some P1, and it is vertical in twos that scanning plane direction, scan objects plane (for example vertical wall) and pushing away swept direction (garage's working direction); The sampling time of laser scanner ignores, then
d = ∫ t 1 t 2 V ( t ) dt - - - ( 1 )
Wherein, In the process that look-up table is set up; Can make camera and laser scanner on the hoistable platform of precision control, move, this moment, speed was very even, and V (t) is constant; They are constant for the difference of the sampling time between the same impact point, can accurately set up the mapping between the data of laser scanner and camera on the two direction of motion through this mistiming.
But when ordinary production; The speed that vehicular platform moves ahead is uneven often; Though with the spacing of wall target intersection fix between laser scanner and the camera this moment; But speed is unfixing, and the mistiming between laser scanner and the camera is just unfixing yet, so relative complex is wanted in the foundation of mapping relations.For this reason; Solution of the present invention is to utilize gps data; Obtain camera and the two speed V (t) on direction of motion of laser scanner that rigidity is bound in real time; Try to achieve sensor through speed V (t) according to formula (1) and move the required mistiming t2-t1 of this fixed range, set up the laser scanning line of garage's direction and the correspondence of camera thus and push away the mapping of pipeline purging, compose colored process thereby accomplish three-dimensional point cloud.
Be elaborated for a preferred embodiment of the parallel binding of camera among the present invention and laser scanner to Fig. 5 below in conjunction with Fig. 3.
Fig. 3 is the parallel binding method schematic flow sheet with the rigidity of laser scanner of line-scan digital camera of the present invention.The purpose of the parallel binding of rigidity is in order to be easy to the processing of line-scan digital camera and laser scanner late time data, therefore preferably its relative line orientation and attitude to be adjusted into as far as possible near zero.
As shown in Figure 3, the parallel binding of concrete rigidity may further comprise the steps: step 11, with the center-aligned of the front nodal point and the laser scanner of line-scan digital camera; Step 12, the angle of adjustment laser scanner and camera makes the two " imaging line " parallel; Step 13, the subtended angle between adjustment laser scanner and the camera makes the constant distance between the two " imaging line "; Step 14 is carried out the rigidity binding with line-scan digital camera that regulates and laser scanner.
Owing to handle for ease of late time data; Need relative line orientation and the attitude of line-scan digital camera and laser scanner to be adjusted into as far as possible near zero, therefore, in step 11; The alignment at the two center can be considered does preliminary rough preparation, accurately adjusts through later step again.Because in the practical operation, though the front nodal point of line-scan digital camera and can be regarded as abstract particle with the center of laser scanner, two instruments itself have certain volume and shape, for example have the profile of rectangular parallelepiped.Therefore; Described " front nodal point of line-scan digital camera and the center-aligned of laser scanner " generally can be regarded as; Will be as far as possible point-blank at the center of the node of installing camera and laser, and this straight line is substantially perpendicular to the scanning plane of laser and the scanning plane of camera.And the front nodal point of camera roughly is positioned on the light cone of laser.Because if the front nodal point of camera can be put on the laser light cone; The RGB of camera assignment on the cloud atlas will become very simple so; As long as the value of the pixel of equal angular is composed the point to the cloud atlas of corresponding angle; Just can obtain color cloud picture, the purpose of eigen is exactly in order to compose colour to cloud atlas a physical basis to be provided.
Below in conjunction with the concrete embodiment that installs the method for parallel binding adjustment is described.For example, can make line-scan digital camera through having the double-deck rebound of three setting nuts, be installed in parallel in the laser scanner summit directly over.
Because the mounting plane of camera on laser can absolute parallel, causes the two " imaging line " not parallel yet.With scan objects is that metope is an example, and the intersection of laser scanning plane and wall target and the intersection of line-scan digital camera image synchronization and wall promptly can be considered " imaging line " of the two respectively.In order to adjust the depth of parallelism between the two, can use clearance gauge, for example the clearance gauge of minimum 0.02 millimeter resolution according to 3 methods that allocate face, utilizes that porose side of clearance gauge, and plug is in the centre of three setting nuts.Realize the some contact for real, three points will have one 1 millimeter basic clearance gauge in advance, otherwise it is just inaccurate the face contact to occur, and pad that can not the tool using steel, and all easy deformation aluminium, copper are got rusty, and use the tool steel clearance gauge to realize.Wherein installed surface is an example in the length of side 200 of installation direction; If headstock is exposed to the north; Camera is installed in the right side of garage's direction, and then this length of side is that the precision of the north and south length of side adjustment of installed surface is the 0.02/200=0.0001 radian, and the distance of three adjustment points can be located at 200 millimeter.
And; Preferably the angle according to camera and laser scanner calculates direction and the thickness that will adjust; The angle of camera and laser scanner also promptly scans the angle of two straight lines on the wall target; Can carry out synchronous scanning imaging through laser scanner and camera, measure same place from cloud atlas and image and try to achieve.
Please with reference to Fig. 4; Make laser scanner and line-scan digital camera be installed on the controlled hoistable platform of a precision, scanning is supposed at a certain fixed time from lower to upper; The intersection of laser and line-scan digital camera and metope is respectively intersection Y and intersection R, and then the installation angle of laser scanner and camera promptly is defined as: Δ R i=alpha-beta, the thickness of clearance gauge are ds, and the thickness that then need adjust is ds * Δ R iWherein, α and β can be by computes:
α=Δt L×V/d;
β=Δt C×V/d;
Wherein, Δ t LBe the mistiming of laser scanning line through some point1 and some point2; Δ t CBe the mistiming of camera sweep trace through some point1 and some point2; V is the platform ascending velocity, and for example, in one embodiment, the measured value of V is 0.007608652m/s; D is the distance between a point1 and the point2, i.e. the length of straight-line segment B between these 2.
In order to confirm monumented point point1 and point2; And the scanning of line-scan digital camera and laser scanner " imaging line " Y, R; Can select for use orthogonal traffic reflective marker as target; The center is 2 centimetres squares a common paper, and target is arranged whole visual fields that then will as far as possible be full of the line-scan digital camera horizontal direction.
About the precision of angle adjustment, the operating distance of this angle should be the threshold of the full field angle of camera at object space, allows the limit difference can be made as probably about 1% degree.This numerical value means when camera during apart from 60 meters of targets; Field of view angle is 90 degree, and the true field width is 120 meters, then differs from the 1% object space error calculated according to limit and should be 120*0.01/57=0.02 rice; Pixel of less than this moment; If the camera sighting distance is far away, it is clear to focus, and this limit difference does not just seem harsh.
Next, the subtended angle between step 13 adjustment laser scanner and the camera is described, the generation of subtended angle is from laser scanner and line-scan digital camera installation errors equally.
Make between " imaging line " parallelly owing to only adjust angle, still can not guarantee parallel between camera scanning plane and the laser scanning face, therefore need the subtended angle of further adjustment between the two.The purpose of said subtended angle adjustment is that the camera scanning plane is parallel with laser scanning face, and this two parallel plane is perpendicular to the line between laser scanner center and the camera center.
Be example still with the mounting means in the above-mentioned angle adjustment; Because the center (front nodal point) of laser scanner and camera can not be installed in one above the particle; So just have one apart from r installing between them; If greater than the distance between camera front nodal point and the laser summit, can judge that two scanning planes are not parallel this moment to the distance between the intersection of the intersection of laser scanning plane and wall target and line-scan digital camera image synchronization and wall less than perhaps.Can be through adjusting another angle that laser and line-scan digital camera are installed; Make laser scanner and line-scan digital camera separately and the distance between the intersection of wall target remain constant; Identical with the two centre distance r; And do not change with the distance of wall target from sensor, the angle that will will adjust this moment among the present invention is called subtended angle.The method of adjustment and the angle of subtended angle are similar; Also need measure same place from cloud atlas and image; Calculate distance between two parallel intersections and whether equal laser scanner and line-scan digital camera, re-use clearance gauge according to the subtended angle size of calculating and regulate the setting angle of camera on laser in the position distance of installing.
Below in conjunction with Fig. 5 the calculating of subtended angle is described.Subtended angle is to calculate on the basis after angle mixes up.Generally need reset some targets point as a token of; Measure the time at this monumented point center respectively from a cloud; Measure the line number at monumented point center from the image of camera, the frequency according to line-scan digital camera triggers calculates the time of monumented point on image; Calculate the time difference of each monumented point camera and laser scanner; Difference is calculated the distance of laser and camera formed intersection when scanning metope according to the velograph of hoistable platform thus, if this distance is not equal to laser scanner central point and the camera front nodal point distance h on the vertical direction when vertical installation, for example; Be 170mm in an embodiment of the present invention, then according between difference just can calculate the angle that will adjust and the thickness of clearance gauge.
Its computing method are that the perpendicular bisector of at first obtaining the monumented point line arrives the vertical apart from L of projection centre, obtains the elevation Z of the corresponding laser scanner of each monumented point then CiWith camera elevation Z Ji(being the two height that is moving with platform), two elevations are subtracted each other and are subtracted the adjustment component vi that camera front nodal point and the distance h (is example with 170mm) on laser scanner summit obtain corresponding each monumented point again,
v i = Z ci - Z ji - h L
The thickness that need adjust the then average value mu of available all monumented points is tried to achieve, and formula is following:
μ = 1 N Σ 1 N v i
Need the thickness=2 * ds * μ mm of adjustment, wherein, ds is a clearance gauge thickness.With clearance gauge thickness is that 230mm is an example, if the thickness of the need that calculate adjustment has not just had necessity of adjustment less than 0.02 millimeter.
The target of subtended angle adjustment can be selected orthogonal traffic reflective marker equally for use, and the center is 2 centimetres squares a common paper, and target is arranged whole visual fields that will as far as possible be full of the line-scan digital camera horizontal direction.
The parking stall of this adjustment ornaments and monumented point lay with the subtended angle adjustment in the ring-type cloud atlas time not quite alike, require here:
1) monumented point is basically on a horizontal line;
2) projection centre is positioned on the perpendicular bisector of monumented point line;
3) projection centre from monumented point how far by the decision of line-scan digital camera field angle require the monumented point point symmetry cover within the visual field, for example field angle is 70 to spend among Fig. 5;
4) 84 coordinates of monumented point also require to measure and;
The scan mode of subtended angle adjustment is the same with the process of angle adjustment, also will measure the laser center position when beginning to scan and finishing to scan; Camera adopts external trigger, triggers camera (100 samplings of suggestion per second) by laser.
The adjustment of subtended angle is based upon the next item up, on the basis that angle has been adjusted, allows the limit difference probably about 1% degree.With regard to the parameter of the foregoing description, after this just meaned that adjustment finishes, the spacing that subtended angle produced also had 1 centimetre error at most about 20 meters; 1 centimetre, according to the speed of hoistable platform, in vertical direction; Be equivalent to carry out the error of 1 pixel on the image of vertical-horizontal proportion chi adjustment back; If the camera sighting distance is far away, it is clear to focus, and this limit difference does not just seem harsh.
Be stressed that present embodiment only as a kind of preferred implementation that realizes parallel binding, makes the method that otherwise realizes parallel binding, only otherwise break away from spirit of the present invention and category, equally can fall into claim scope of the present invention.For example, in the experimental situation of strictness control, be designed with the target metope of horizontal parallel linear mark; Perhaps linear array CCD camera is demarcated etc. with the parameter calibration method of prior art.Though these methods realize going up the complicacy more next than the preferred embodiment, also can realize the object of the invention.
The above description of this invention is illustrative, and nonrestrictive, and those skilled in the art is understood, and within spirit that claim limits and scope, can carry out many modifications, variation or equivalence to it, but they will fall in protection scope of the present invention all.

Claims (9)

1. a method that generates laser color cloud picture is provided with a plurality of impact points in scene to be measured, it is characterized in that, may further comprise the steps:
S10 carries out the parallel binding of rigidity with line-scan digital camera and laser scanner, so that the scanning plane of the two is parallel, and two parallel scanning planes are perpendicular to the line at camera front nodal point and laser scanner center;
S20 makes line-scan digital camera and laser scanner carry out synchronous scanning;
S30 sets up laser scanning angle and its corresponding relation between the pixel sequence number on the linear array image of impact point of the same name, and this mapping relations are referred to as color lookup table;
S40 sets up each laser point and its RGB information corresponding relationship in the laser point cloud according to difference sweep time of color lookup table and laser scanner and line-scan digital camera, wherein:
Among the said step S10, it is may further comprise the steps that line-scan digital camera and laser scanner are carried out the parallel binding of rigidity:
S11 is with the center-aligned of the front nodal point and the laser scanner of line-scan digital camera;
S12, the angle of adjustment laser scanner and line-scan digital camera makes the imaging line of the two parallel;
S13, the subtended angle of adjustment laser scanner and line-scan digital camera makes the constant distance between the two the imaging line;
S14 carries out the rigidity binding with line-scan digital camera that regulates and laser scanner.
2. the method for generation laser color cloud picture according to claim 1; It is characterized in that; Among the said step S10, be with line-scan digital camera through having the double-deck rebound of three setting nuts, be installed in parallel in the laser scanner summit directly over; And between said setting nut, be provided with clearance gauge, carry out the parallel adjusting of the two through regulating clearance gauge.
3. the method for generation laser color cloud picture according to claim 1; It is characterized in that; The method of the angle of adjustment laser scanner and line-scan digital camera is: make laser scanner and line-scan digital camera at the uniform velocity mobile synchronously with speed V, a plurality of impact points that are arranged at the ring-type metope are carried out to picture; Calculate laser scanning line and line-scan digital camera sweep trace respectively through the mistiming Δ t of the some point1 in the impact point with some point2 L, Δ t C, the clearance gauge distance that then need adjust is: ds * Δ R i, wherein, Δ R i=alpha-beta, α=Δ t L* V/d, β=Δ t C* V/d, d are a point1 and the air line distance of putting point2, and ds is a clearance gauge thickness.
4. the method for generation laser color cloud picture according to claim 1; It is characterized in that; The method of the subtended angle of adjustment laser scanner and line-scan digital camera is: make laser scanner and line-scan digital camera at the uniform velocity mobile synchronously with speed V, a plurality of impact points that are arranged at metope are carried out to picture; The perpendicular bisector of obtaining the monumented point line arrives the vertical apart from L of projection centre, obtains the elevation Z of the corresponding laser scanner of each monumented point then CiWith line-scan digital camera elevation Z Ji, said elevation is the two distance apart from starting point when monumented point is sampled, two elevations subtract each other and subtract the adjustment component v that line-scan digital camera front nodal point and the distance h on laser scanner summit obtain corresponding each monumented point again i,
Figure FSB00000665329700021
needs the thickness=2 * ds * μ of adjustment; Wherein, ds is a clearance gauge thickness;
Figure FSB00000665329700022
N is a monumented point number in the image.
5. the method for generation laser color cloud picture according to claim 1 is characterized in that, the parallel binding of said rigidity is to be fixed together with the scanning plane and the line-scan digital camera of mechanical mode with laser scanner, makes it can not relative motion; Perhaps laser scanner and line-scan digital camera are carried out synchronous follow-up control with the mode of mechanical/electrical.
6. the method for generation laser color cloud picture according to claim 1 is characterized in that, among the said step S20, line-scan digital camera and laser scanner carry out synchronous scanning to be realized through the two being fixed on the controlled hoistable platform of a precision.
7. the method for generation laser color cloud picture according to claim 1 is characterized in that, among the said step S30, the foundation of color lookup table is according to as a token of identical point coordinates and the coupling of image information on the laser point cloud; At first; In a cloud, identify monumented point; Find the angle of corresponding point in the corresponding polar coordinates file afterwards according to gps time information subsidiary in the coordinate of monumented point, in image, measure the row number of this monumented point simultaneously, i.e. which pixel in its corresponding line-scan digital camera; Thereby set up the mapping relations between the corresponding polar angle of pixel of line-scan digital camera with laser scanner.
8. the method for generation laser color cloud picture according to claim 1 is characterized in that, laser scanning angle and its corresponding relation between the pixel sequence number on the linear array image are non-linear in the said color lookup table.
9. the method for generation laser color cloud picture according to claim 1; It is characterized in that; Among the said step S40; Utilize gps data; Obtain line-scan digital camera and the two speed V (t) on direction of motion of laser scanner that rigidity is bound in real time; Try to achieve line-scan digital camera and the required mistiming t2-t1 of laser scanner displacement d through speed V (t) according to formula
Figure FSB00000665329700023
, the correspondence of setting up laser scanning line and line-scan digital camera thus pushes away the mapping of pipeline purging, and wherein d is the distance between line-scan digital camera front nodal point and the laser scanner center.
CN2010102372655A 2010-07-22 2010-07-22 Method for generating laser color cloud picture Active CN101901501B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2010102372655A CN101901501B (en) 2010-07-22 2010-07-22 Method for generating laser color cloud picture

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2010102372655A CN101901501B (en) 2010-07-22 2010-07-22 Method for generating laser color cloud picture

Publications (2)

Publication Number Publication Date
CN101901501A CN101901501A (en) 2010-12-01
CN101901501B true CN101901501B (en) 2012-02-08

Family

ID=43227007

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2010102372655A Active CN101901501B (en) 2010-07-22 2010-07-22 Method for generating laser color cloud picture

Country Status (1)

Country Link
CN (1) CN101901501B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8731247B2 (en) 2012-01-20 2014-05-20 Geodigital International Inc. Densifying and colorizing point cloud representation of physical surface using image data

Families Citing this family (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103186819A (en) * 2011-12-28 2013-07-03 象山县供电局 Quadratic optimization method for electric power facility site selection based on three-dimensional geographic information system
CN102798380B (en) * 2012-07-09 2014-11-26 中国人民解放军国防科学技术大学 Method for measuring motion parameters of target in linear array image
CN103606151B (en) * 2013-11-15 2016-05-04 南京师范大学 Based on the virtual geographical scene method for auto constructing on a large scale of imaging point cloud
CN103644896B (en) * 2013-12-30 2015-11-04 中国电建集团成都勘测设计研究院有限公司 A kind of engineering geological mapping method based on 3 D laser scanning
CN105761301B (en) * 2014-12-16 2018-12-04 财团法人金属工业研究发展中心 Three-dimensional laser scanning system and its colour point clouds image method for building up
CN104680587B (en) * 2014-12-23 2018-01-16 山东丽鹏股份有限公司 A kind of laser engraving machine five-axis linkage processing method
CN105551068B (en) * 2015-12-07 2018-07-24 中国人民解放军空军装备研究院雷达与电子对抗研究所 A kind of synthetic method of 3 D laser scanning and optical photograph
CN107830800B (en) * 2017-10-26 2019-11-12 首都师范大学 A method of fine elevation is generated based on vehicle-mounted scanning system
CN107991662B (en) * 2017-12-06 2023-12-15 南京添马机械科技股份有限公司 3D laser and 2D imaging synchronous scanning device and scanning method thereof
CN110308433B (en) * 2018-03-27 2023-03-17 深圳砺剑天眼科技有限公司 POS data and image triggering time matching method of laser radar system
CN110148180B (en) * 2019-04-22 2021-06-08 河海大学 Laser radar and camera fusion device and calibration method
CN110533686B (en) * 2019-10-08 2022-03-04 凌云光技术股份有限公司 Method and system for judging whether line frequency of linear array camera is matched with object motion speed
CN110717987B (en) * 2019-10-11 2023-03-14 徐州中矿奥特麦科技有限公司 Data fusion method based on stock ground model and temperature model
CN110969701B (en) * 2019-11-29 2023-12-29 司空定制家居科技有限公司 Method, device and storage medium for modeling house
CN113960046A (en) * 2021-10-08 2022-01-21 深圳市霄光科技有限公司 Surface detection method and system
CN114494550B (en) * 2021-12-30 2022-11-22 北京城市网邻信息技术有限公司 WebGPU-based rendering method, electronic device and storage medium

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1374620A (en) * 2002-03-27 2002-10-16 深圳市特得维技术有限公司 Computerized 3D visual color scan system and its scanning mode
CN101581575A (en) * 2009-06-19 2009-11-18 南昌航空大学 Three-dimensional rebuilding method based on laser and camera data fusion
CN101697233A (en) * 2009-10-16 2010-04-21 长春理工大学 Structured light-based three-dimensional object surface reconstruction method

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4038726B2 (en) * 2003-09-03 2008-01-30 株式会社日立プラントテクノロジー Image matching method
KR100800554B1 (en) * 2007-07-03 2008-02-04 (주)지에스엠솔루션 Texture mapping method of 3d feature model using the camera and laser scanner

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1374620A (en) * 2002-03-27 2002-10-16 深圳市特得维技术有限公司 Computerized 3D visual color scan system and its scanning mode
CN101581575A (en) * 2009-06-19 2009-11-18 南昌航空大学 Three-dimensional rebuilding method based on laser and camera data fusion
CN101697233A (en) * 2009-10-16 2010-04-21 长春理工大学 Structured light-based three-dimensional object surface reconstruction method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
杨勇,钟若飞,康永伟,秦涛.车载激光与单目线阵相机的数据融合.《首都师范大学学报( 自然科学版)》.2010,第31卷(第2期),全文. *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8731247B2 (en) 2012-01-20 2014-05-20 Geodigital International Inc. Densifying and colorizing point cloud representation of physical surface using image data
US9053572B2 (en) 2012-01-20 2015-06-09 Geodigital International Inc. Densifying and colorizing point cloud representation of physical surface using image data
US9269188B2 (en) 2012-01-20 2016-02-23 Geodigital International Inc. Densifying and colorizing point cloud representation of physical surface using image data

Also Published As

Publication number Publication date
CN101901501A (en) 2010-12-01

Similar Documents

Publication Publication Date Title
CN101901501B (en) Method for generating laser color cloud picture
CN203084734U (en) System for regenerating virtual object
CN106056587B (en) Full view line laser structured light three-dimensional imaging caliberating device and method
CN102003938B (en) Thermal state on-site detection method for large high-temperature forging
CN103292710B (en) A kind of distance measurement method applying binocular vision vision range finding principle
CN102364299B (en) Calibration technology for multiple structured light projected three-dimensional profile measuring heads
CN100458359C (en) Small-displacement measuring system in long-distance plane
CN108759669B (en) Indoor self-positioning three-dimensional scanning method and system
CN110849268B (en) Quick phase-height mapping calibration method
CN104851088B (en) The line-scan digital camera calibration facility and scaling method of a kind of high ferro detecting system
CN105115560B (en) A kind of non-contact measurement method of cabin volume of compartment
CN109269466A (en) Target surface relative pose measurement method and system based on characteristic point
CN101509763A (en) Single order high precision large-sized object three-dimensional digitized measurement system and measurement method thereof
Rumpler et al. Automated end-to-end workflow for precise and geo-accurate reconstructions using fiducial markers
CN103759669A (en) Monocular vision measuring method for large parts
CN109141226A (en) The spatial point coordinate measuring method of one camera multi-angle
CN100343625C (en) Measuring splice method and device of large three-dimensional body shape based on splicing target
CN104567727A (en) Three-dimensional target and global unified calibration method for linear structured light profile sensor
CN108198224A (en) A kind of line-scan digital camera caliberating device and scaling method for stereo-visiuon measurement
KR20110027522A (en) The method for geometric registration for images data and lidar data and the apparatus thereof
CN101487702A (en) Binocular vision based traffic accident on-site photogrammetric survey method
CN108180888A (en) A kind of distance detection method based on rotating pick-up head
CN109920009A (en) Control point detection and management method and device based on two dimensional code mark
CN105043304A (en) Novel calibration plate and calibration method for performing length measurement by using calibration plate
CN1605830A (en) Large-scale three dimensional shape and appearance measuring and splicing method without being based on adhesive mark

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant