CN103049622A - Reverse engineering processing system and method based on space code projection - Google Patents
Reverse engineering processing system and method based on space code projection Download PDFInfo
- Publication number
- CN103049622A CN103049622A CN2013100157427A CN201310015742A CN103049622A CN 103049622 A CN103049622 A CN 103049622A CN 2013100157427 A CN2013100157427 A CN 2013100157427A CN 201310015742 A CN201310015742 A CN 201310015742A CN 103049622 A CN103049622 A CN 103049622A
- Authority
- CN
- China
- Prior art keywords
- point
- processing
- reverse
- curved surface
- cloud
- 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.)
- Pending
Links
Images
Landscapes
- Length Measuring Devices By Optical Means (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a reverse engineering processing system and a method based on space code projection. The reverse engineering processing system comprises a collection device, a treatment device and a processing device, wherein the collection device is connected with the treatment device which is connected with the processing device. The processing method comprises the following steps, 1) projecting space codes on objects being collected and obtaining spacial size coordinate information of a measured object through video; 2) treating and reconstructing collected images to obtain an optimized curve surface module and 3) analyzing and calculating the obtained curve surface module to form a tool path. The reverse engineering processing system and the method based on the space code projection is good in robustness and adaptivity, obtains the spacial size coordinate information of the measured object through video, reconstructs curve surface module, and accordingly the tool path is formed and conversed into controlling instructions and rapid molding on a controlling device is achieved. The reverse engineering processing method combines a visual measurement technology and a rapid processing technology to build an automatic measurement and molding system for the processing of high precision equipment.
Description
Technical field
The present invention relates to the reverse-engineering based on the space encoding projection, design specifically a kind of reverse-engineering system of processing and method based on the space encoding projection.
Background technology
The research of reverse-engineering and application, the domestic development time is evening relatively.The middle and later periods nineties in last century, reverse-engineering just causes domestic scholar's concern and attention.
The Curve Reconstruction based on cross-section data that the research group of Nanjing Aero-Space University proposes does not have to use the method based on feature, namely for B-spline curves of data point overall fit in cross section, then carries out covering based on the B-spline curves group and calculates.The Li Gang of research group of Shandong University is according to ultimate principle and the flow process of reverse-engineering, existing measuring equipment, reverse software, CAD modeling software etc. are organically combined, the product mock-up is carried out three-dimensional data measurement, data processing, curve reestablishing of profile etc., and on this basis, the principle features of contact and contactless measurement and suitable occasion have been studied; How studied and guaranteeing measuring accuracy, correctly reflecting under the condition of geometric configuration, optimized measuring route, reduced measurement data points, improved and measure efficient, solved the variety of issue that runs in measuring.The Wu of Zhejiang University keeps combination is thought in spring transportation with theory and practice method, characteristic line to cloud data and grid data extracts, the crestal line of triangle gridding extracts and the shape recognition of cloud data has been carried out certain research, seek counting yield high, the computerized algorithm that the result is attractive in appearance.Tan of Zhejiang University brilliance carries out theory study and applied research to the Shape Editing problem in a cloud and the grid model, be intended to improve the reusability of existing model and distortion editor's efficient, reduce the workload of three-dimensional modeling and animation, proposed a method based on the extensive some cloud curved surface Free Transform of ObbGraph.The method combines traditional Free Transform (Free Form Deform) advantage that feature keeps in advantage and the Differential Mesh deformation technology more fast, thereby when improving deformation efficiency, the local geometric details of holding point cloud curved surface preferably.And definition and the closed solution thereof of vertex deformation gradient on the some cloud curved surface have been provided.Be different from traditional deformation gradient that is defined on the tri patch, the vertex deformation gradient of some cloud curved surface directly is defined in each summit of a cloud curved surface, thereby need not the topological structure of similar grid.Spline surface match and the problem such as grid model deformation technology of what army of Shandong University on the grid carried out further investigated, for batten Fitting on the grid, proposes a kind of new method of constructing reasonable bicubic surface at irregular control mesh.By the decomposition of even cubic B spline base function and the classification of subbase function, use the method constructing curve on irregular control mesh that is similar to B-spline surface.Simultaneously, for grid model adjusted size problem, propose a kind of method for adjusting of the new maintenance aspect of model, can well keep the feature of the perception sensitizing range of model, will be out of shape and be distributed to de-militarized zone as far as possible.For the grid model problem on deformation, the model deformation method that proposes a kind of feature-sensitive is paid close attention to the feature details of grid model more.Specifically by the thought of the responsive tolerance of use characteristic in dual mess, directly normal vector is taken into account, so the Laplace operator of feature-sensitive is very responsive to the feature details.
Because domestic manufacturing develops rapidly, emerge large quantities of high-tech companies and promoted reverse Engineering Technology and develop rapidly in China.Xi'an High-Tech AEH Industrial Metrology Co., Ltd., it has independent intellectual property right in core technologies such as digital control system, Survey Software and optical image information treatment technologies, reaches advanced world standards in CNC control under the foundation of three-dimensional data processing model, PC bus, Computer Image Processing and measurement of coordinates technical research field.RE-SOFT software by Zhejiang University and the joint development of Hangzhou one tender tea leaves software company, the Reverse modeling strategy that provides the Reverse modeling normal form of a new generation take multi-stage characteristics extraction, evaluation, design, computing, reproduction and analysis as core to show unique characteristics will be put the most advanced method for solving that cloud is converted into the senior information such as design concept, method for designing, the principle of work and power, Engineering constraint and aesthetic theory for the user provides.The Development of Modular technology has been used in software development, is comprised of data exchange module, reverse moulding module, Analysis and measurement module, aid module, figure demonstration and user interactive module.
Reverse-engineering has obtained the progress of highly significant at the aspects such as exploitation of data processing, patch match, rule feature identification, special-purpose business software and spatial digitizer.But in actual applications, whole process still needs a large amount of man-machine interactivelies, operator's experience and quality influence the quality of product, the fairness of automatic Reconstruction curved surface is difficult to guarantee, and not with the research and development case of reality processing combination, therefore solve the problems of the technologies described above in the urgent need to a kind of new method and system.
Summary of the invention
The present invention is just for the technical matters that exists in the prior art, a kind of reverse-engineering system of processing and method based on the space encoding projection is provided, the method is obtained the coordinate information of testee space scale by video means, convert thereof into and control instruction, rapid shaping on program-controlled equipment, vision detection technology is mutually integrated with the rapid processing technology, set up one towards automatic detection and the formation system of high-precision equipment processing.
In order to achieve the above object, technical scheme of the present invention is as follows, a kind of reverse-engineering job operation based on the space encoding projection, it is characterized in that, said method comprising the steps of 1) to being gathered object projecting space coding, obtain again the coordinate information of the space scale of testee by video means, 2) image that collects is processed, reconstruct also is optimized for surface model; 3) surface model that generates is carried out analytical calculation, generate cutter path.
As a kind of improvement of the present invention, in the described step 2 image being carried out pretreated process comprises, point cloud denoising, point cloud compressing, hole repair and the distortion of some cloud, in the three dimensional point cloud always with the existence of noise spot, about 0.1%-5% noise spot in the cloud data that the statistics gained is measured.Noise spot needed cloud data is carried out denoising before model reconstruction so that the model curved surface of reconstruct is rough, the situation of appearance distortion.For three-dimensional point cloud, the reason that noise produces is from two aspects, and the one, there is defective testee in itself, such as rough surface, ripple etc. is arranged, is the equipment of measuring system on the other hand, such as the nonlinearity erron of ccd video camera, the electrical noise of system etc.
Described some cloud denoising is by rejecting the singular point noise in the k neighborhood and utilizing the average drifting iteration to carry out offset point and adjust two steps combinations and finish.Wherein the singular point noise is in the noise spot data, and some point is obviously and is in different range on every side, is presented as burr in model, therefore utilizes first the interior filtering algorithm of k neighborhood directly to reject for singular point; Described adjustment offset point refers to depart from entopic noise spot for some, need to carry out the position adjustment, makes it more near the position of original curved surface, and therefore recycling the average drifting iterative algorithm carries out the offset point adjustment.
Described point cloud compressing is on the Voronoi figure of cloud data structure, utilizes fast marching algorithms constantly to expand sample area, utilizes simultaneously the solstics sampling condition to filter out qualified sampled point.
Described hole is repaired, comprise that mainly (1) extracts hole edge feature point (2) and utilizes radial basis function to simulate curved surface (3) on the hole according to certain step-length according to the information of hole peripheral point, sample at curved surface (4) adopt Moving Least to adjust to resulting point so that the curved surface after repairing can be connected with curved surface is smooth on every side.
Described point cloud model distortion is optimized, that point cloud model is carried out Free Transform more to be adapted to the data model of processing, the characteristic of described Free Transform mainly is that the character by basis function shows, and introduces different basis functions, just can form the Free Transform of different qualities.
As a kind of improvement of the present invention, described step 3 comprises the cross section distance computation, and cutter location calculates, and the cutter spacing path generates, and numerical control code generates.
A kind of reverse-engineering system of processing based on the space encoding projection, it is characterized in that, described system comprises harvester, treating apparatus and processing unit (plant), described harvester connects treating apparatus, described treating apparatus connects processing unit (plant), and described harvester, treating apparatus, processing unit (plant) device adopt serial mode to connect.
As a kind of improvement of the present invention, described harvester comprises camera, projector, capture card, described treating apparatus is embedded main board, described mainboard comprises pci interface chip, arm processor, RAM, the level transferring chip that CPLD consists of, described treating apparatus is numerically-controlled machine, the testee image is identical by two intrinsic parameters, be the CCD camera of symmetrical contour placement by synchronous recording, by the video frequency collection card collection, video frequency collection card is connected with system board by pci bus, the ARM chip is controlled pci interface chip by interrupt mode, its exchanges data cushions by dual port RAM, described ARM obtains the laggard line algorithm of image information by pci interface and processes, and the control routine of final generating three-dimensional lathe exports control routine to processing unit (plant) by USB interface.
As a kind of improvement of the present invention, described projector adopts the LCD liquid crystal imaging, projector distance is set to 1.2 5.7 meters, liquid crystal board is of a size of 0.7 inch, the standard display resolution is 1024 * 768 pixels, nominal contrast 450: 1, nominal light 2000 lumens, camera lens F value 1.8 2.15,23 27.6 millimeters of focal lengths.
As a kind of improvement of the present invention, described camera adopts the CCD camera, adopt 1/3 inch digital imagery plate, it adopts 1/3 inch digital imagery plate, 440,000 pixels, the horizontal resolution of 535 line high-resolutions, minimal illumination 0.7 lumen, signal to noise ratio (S/N ratio) is 50dB, and built-in 10 bit digital processors are supported automatic backlight compensation and Automatic white balance.
With respect to prior art, the present invention's design has good robustness and adaptivity, has widely three-dimensional acquisition detection system, obtain the coordinate information of testee space scale by video means, convert thereof into and control instruction, rapid shaping on program-controlled equipment, vision detection technology is mutually integrated with the rapid processing technology, set up one towards automatic detection and the formation system of high-precision equipment processing.(revise the same with letter of authorization).
Description of drawings
Fig. 1 is job operation overall flow figure of the present invention;
Fig. 2 is system of processing one-piece construction schematic diagram of the present invention.
Wherein 1 is camera, and 2 is projector, and 3 is image pick-up card, 4 is pci interface, and 5 is pci interface chip, and 6 is arm processor, 7 is lathe, and 8 is RAM, 9 level transferring chip for the CPLD formation, 10 is keyboard, 11 is LCD display, and 12 is FLASH, and 13 is D/A converter, 14 is the VGA interface, and 15 is USB interface.
Embodiment
In order to deepen the understanding of the present invention and understanding, the present invention will be described in detail and introduction below in conjunction with the drawings and specific embodiments.
Embodiment 1:
Referring to Fig. 1, a kind of reverse-engineering job operation based on the space encoding projection, said method comprising the steps of, 1) to being gathered object projecting space coding, obtain again the coordinate information of the space scale of testee by video means, 2) image that collects is processed, reconstruct also is optimized for surface model; 3) surface model that generates is carried out analytical calculation, generate cutter path.
In the step 1, the mathematical model of optical projection system, in structure light projection system, based on the light path principle of reversibility, preferred view instrument 2 can be considered a reverse desirable pinhole camera, the projection matrix P of projector 2 can be decomposed into 3 * 3 Intrinsic Matrix Ap, 3 * 3 rotation matrix Rp, the translation vector Tp between 3 * 1 coordinate system.Set up the mathematical model of preferred view instrument based on the pin-hole model of camera 1; Based on the coding method on striped border, adopt first the method for graph theory, by the centronucleus of gained coding, i.e. center striped and near coding thereof manually are set.These codings have good symmetry and saltus step, are convenient to digital picture and detect and extract.Then utilize the Hadamard algorithm to generate fast an initial encoder matrix.Recycling Commutative Algebra mode is changed the different lines in the matrix, so that initial several row of this matrix satisfy the designing requirement of striped boundary coding, this advantage is to travel through whole space encoder, thereby can find preferably coded sequence.
In the step 2, pretreated process comprises, the denoising of some cloud, and point cloud compressing, hole repair and cloud data distortion are optimized
,Described some cloud denoising be, in the three dimensional point cloud always with the existence of noise spot, about 0.1%-5% noise spot in the cloud data that the statistics gained is measured.Noise spot needed cloud data is carried out denoising before model reconstruction so that the model curved surface of reconstruct is rough, the situation of appearance distortion.For three-dimensional point cloud, the reason that noise produces is from two aspects, and the one, there is defective testee in itself, such as rough surface, ripple etc. is arranged, is the equipment of measuring system on the other hand, such as the nonlinearity erron of ccd video camera, the electrical noise of system etc.
Algorithm of the present invention is divided into rejecting the singular point noise in the k neighborhood and utilizing the average drifting iteration to carry out offset point in general adjusts this two steps, reject singular point, in the noise spot data, some point is obviously and on every side to be in different range, in model, be presented as burr, directly reject for singular point.Adjust offset point, depart from entopic noise spot for some, need to carry out the position adjustment, make it more near the position of original curved surface.
Described point cloud compressing is that unnecessary cloud data can affect three-dimensional reconstruction efficient.Greatly alleviated the burden of computer data storage and computing after cloud data simplified.Fast Marching Method is Level Set Method, is used for pursuit movement border traveling process.The border here refers to connect the critical region in inside and outside zone, and critical region is closed curve in two dimensional surface, is occluding surface in the three dimensions.On the Voronoi figure of cloud data structure, utilize fast marching algorithms constantly to expand sample area, utilize simultaneously the solstics sampling condition to filter out qualified sampled point.
It is whether can connect into closed curve according to frontier point and dot cloud hole can be divided into cap holes and non-cap holes two classes that described hole is repaired.Whether cap holes is according to having solid dot to be divided into single hole and island hole.Whether non-cap holes contains by near point its hole that certain feature is divided into the feature hole and without the feature hole.The key step of algorithm: at first extract the hole edge feature point, utilize radial basis function to simulate curved surface on the hole according to the information of hole peripheral point subsequently, subsequently according to certain step-length, sample at curved surface, the point of gained carries out Moving Least resulting point is adjusted, so that the curved surface after repairing can be connected with curved surface is smooth on every side.
Described point cloud model distortion is optimized, that point cloud model is carried out Free Transform more to be adapted to the data model of processing, the characteristic of described Free Transform mainly is that the character by basis function shows, and introduces different basis functions, just can form the Free Transform of different qualities.
The present invention is directed to set forth above some to have designed detailed algorithm as follows:
(1) the present invention adopts the cloud data of KD storage of data structure space encoding projection acquisition gained, finishes the detection of frontier point and being connected of boundary line based on the k neighborhood search algorithm of KD tree.
(2) neighborhood of the data of point cloud boundary point is analyzed, extracted edge feature point, utilize normal vector to choose suitable neighborhood scope, the scope of dwindling sampled point is carried out parametrization to point range simultaneously, for subsequent calculations is simplified.
(3) the surface fitting stage, the kernel function of basis function is radially selected, chosen the kernel function of suitable radial basis function, and on this basis the hole border is expanded, the neighborhood information that takes full advantage of edge feature point in the process of expansion produces filling point, finishes the repairing to hole.
(4) filling point behind the hole repair is adjusted again.Guaranteed that filling point and neighborhood point on every side can be smoothly connected.
As a further improvement on the present invention, the point cloud model behind the hole repair is carried out Free Transform, make it more be adapted to carry out post analysis and processing.The characteristic of described Free Transform mainly is that the character by basis function shows, and introduces different basis functions, just can form the Free Transform of different qualities.Free Transform based on the B spline base function is the typical case of this respect.Distortion has of overall importancely based on the FFD of Bezier, and has locality based on the FFD of B batten.And if Bezier and B batten are combined, then can between of overall importance and locality, find an equilibrium point.Because the point cloud model after the distortion may density diminish, need to carry out a cloud up-sampling simultaneously, Alexa proposes the some cloud up-sampling of a kind of movement-based least square projection.The main thought of the method is to calculate Voronoi figure at moving least squares surfaces because the summit of Voronoi figure be exactly on the curved surface to around had a maximum point of distance, so these summits will be the points of increase.Yet the calculated amount of calculating Voronoi figure at moving least squares surfaces is huge, so Alexa adopts a Local approximation.Above-mentioned algorithm is the some cloud up-sampling that projects on the MLS curved surface, but in the present invention, the some cloud is comprised of triangular plate through behind the surface reconstruction.So must improving top algorithm, the above could be used for this project.Improved step was the 4th step.After the circle of maximum is selected, its center of circle is not to project on the MLS curved surface, put the nearest original point cloud subpoint of middle distance on every side but look for, with the vertical line of this circle left side perpendicular to part plan, the triangle intersect that upwards forms with the three-dimensional point of these three points, what then intersect is exactly the point that newly increases.
Embodiment 2:As a kind of improvement of the present invention, the image after processing is carried out three-dimensionalreconstruction, mainly comprise surface reconstruction and surface smoothing.
Triangulation is surface reconstruction, in the reverse-engineering, surface reconstruction is accepted obtaining and processing of three dimensional point cloud, simultaneously be again follow-up curved space filtering and the basis of fairing, in the reverse-engineering, surface reconstruction is defined as follows: for one group of given measurement data points, curved surface of reconstruct can approach the realistic model surface well.Implementation strategy according to surface reconstruction, the method of present surface reconstruction can be divided into two classes: the one, obtain curve by the measurement data points match and carry out curved surface modeling, another kind is to utilize measurement data points directly to carry out surface fitting, namely be the topological relation of directly from dispersion point cloud, seeking each other, set up curved surface.And research is the most extensive in this process, uses proven technique and is triangulation.
The typical problem that triangulation will solve is: given Scattered data points
, how to set up the connection between each point, so that the set of whole V forms a triangle gridding, and so that mesh quality is higher.Its essence is with triangle gridding and reflects that the neighborhood between the scattered data points concerns, thus the contact of the inward nature between announcement and the expression data point.
At present, for the triangulation under the three-dimensional planar, a kind of solution commonly used is for to do projection, with the problem dimension-reduction treatment on two dimensional surface.And on two dimensional surface, a kind of Triangulation Technique commonly used is the Delaunay triangulation.
For three dimensional point cloud, adopt the BPA spin algorithm to realize triangulation, the main thought of BPA algorithm is: if one is set by the user
Value was not all met other point as the circle of radius before running into three points, these three points just form a triangle so.Since a seed triangle, the BPA algorithm moves (take a leg-of-mutton limit as axis, rotating around this axis such as it) until it meets another point around a limit, forms the another one triangle.Till this process is continued until that all limits all are used as main shaft and rotated.Then from another seed triangle, repeat said process.Until be not taken as the limit of axis.The BPA algorithm is compared to other algorithm, and the memory headroom that needs is relatively less, and very efficient in time, and the effect strong robustness of reconstruct.
For the cloud data of space projection coding gained, the present invention proposes one based on the BPA algorithm of angle constraint.Adopt the BPA algorithm after angle retrains to give up some redundant reference mark, so that the sizing grid that reconstruct obtains reaches unanimity more, be convenient to the processing of follow-up cutter.
Described surface smoothing refers to that the curved surface behind the triangulation can not reach the application requirements of Practical Project, can not directly output to carry out material object processing on the apparatus.The subdivision result being carried out surface smoothing, help so that the more realistic object standard of model has also made things convenient for follow-up path planning design and feed simultaneously, is a very important step.The basic thought of surface smoothing is similar to the filtering in the Digital Image Processing.The basic ideas of filtering are that the pixel value of some points is not only relevant with self value, should be subject to the impact of surrounding pixel value simultaneously.According to certain operator, carry out the filtering of different modes, such as medium filtering, mean filter, gaussian filtering etc.Similar to it, the basic thought of surface smoothing is that the coordinate of each point in the three dimensional point cloud is not only relevant with himself coordinate figure, simultaneously should be subject to the impact of cloud data on every side, angle from the subdivision plane, the point nearer apart from this cloud, its impact should be not less than apart from its point far away.
Owing to reasons such as model itself, measuring accuracy, the errors of calculation, first that there is much noise in curved surface.These noises may not be positioned at single yardstick, the noise that brings such as measuring accuracy and the caused noise that rounds up of calculating are not the same order of magnitude, therefore, surface smoothing under the single scale is processed, or when better eliminating noise, model information is caused damage, or for maintenance model original appearance reduces iterations, thereby suitable abundant not to the small scale noise light.Be directed to this, in the technical program scale factor introduced in the fairing as parameter, with becoming the yardstick Algorithms for Surface Smoothing model is processed, to improve the deficiency under the single scale fairing.And by theoretical proof better fairing effect under the single yardstick under the change yardstick.
Embodiment 3:
As a kind of improvement of the present invention, in the step 3, the model after processing is carried out computational analysis, generate cutter path.The present invention selects to be most widely used uniform cross section cutter path generating algorithm.The method comprises the cross section distance computation, and cutter location calculates, and the cutter spacing path generates, and numerical control code generates.Described cross section distance computation can be calculated with reference to Free-Form Surface Machining feed line-spacing formula, still for the grid surface of complexity, exists details place curvature large, is difficult to problems such as formula calculating, and therefore main the and time considers.Described cutter location calculates, and a kind of method of calculating cutter location is to calculate according to cutter-contact point and curved surface local feature.Because triangle mesh curved surface is difficult to describe with mathematic(al) representation, also just be difficult to according to the cutter-contact point line cutter spacing line of deriving.Therefore the present invention adopts the method for surf offset to calculate cutter location.Curved surface after the skew just becomes the cutter spacing face, tries to achieve the cutter spacing face with plane and surface intersection again.
At first need each the triangle decoupling zero with curved surface, so that each triangle of curved surface is independently, namely when a triangle was moved, remaining triangle can not be affected.If the radius of ball head knife is r, ball head knife and curved surface contact are always r to the centre of sphere of ball head knife distance, and simultaneously because ball head knife remains tangent with curved surface, so the contact is moved r along this leg-of-mutton normal vector in place, contact and just obtained cutter location.Surf offset only needs to get final product according to the direction skew r of each triangulation method vector.But mobile like this rear curved surface is originally to keep like that continuous no longer just.But the set that each independent triangle forms.The place of evagination, triangle is separated from each other, and the local triangle of indent intersects mutually.
Secondly the cutter spacing face after the skew is carried out the plane and ask friendship, calculate the cutter location path-line.For the speed of the computing of asking friendship, set up the Bounding Box tree of curved surface.When like this friendship is asked on each plane, need not travel through each triangle of curved surface.Because what obtain is the line segment of a series of fractures or the line segment of intersection.Therefore need further to process to obtain continuous path-line.Described without interfering the cutter spacing path to generate, method of section calculating path core is to obtain without interfering cutter path.For the part of intersecting lens Fracture, need to carry out interpolation, so that the path can be continuous.And for the part of intersecting, the part that is blocked can produce the path of interference, therefore only need remove the part that is blocked.
The lower problem of equal cross section path efficiency of the present invention proposes a kind of cutter path generation method of five equilibrium aspect.Basic thought etc. aspect is with surface level and surface intersection, can obtain region unit one by one after intersecting, and curved surface those region units on surface level need not be processed, and the region unit of curved surface under surface level will be cut.The path that at first needs generated uniform cross section method.The equal cross section path, water is truncate from top to bottom gets, and obtains the path-line of every one deck.The curve that the path, cross section is positioned under the aspect will be replaced by the straight line on the aspect, and the curve that the path, cross section is positioned on the aspect will directly be given up, like this with regard to self-assembling formation region unit one by one.Path-line for every one deck obtains although be distributed as in form different processing blocks, structurally remains a series of line segment chained list.In order to realize the purpose of block-by-block processing, every layer line segment need to be carried out piece and divide.The advantage of this algorithm is can reduce process time by piece processing on the aspect of grade.
Embodiment 4:
Referring to Fig. 2, a kind of reverse-engineering system of processing based on the space encoding projection, described system comprises harvester, treating apparatus, processing unit (plant), harvester connects treating apparatus, described treating apparatus connects processing unit (plant), and described harvester, treating apparatus, processing unit (plant) adopt serial mode to connect.Described harvester comprises camera 1, projector 2, capture card 3, described treating apparatus is embedded main board, place projector 2 in the middle of the system, encoded by the striped that the striped boundary coding module among the ARM generates, generate the VGA signal by high speed D/A, be input in the projector, carry out the striped coding projection to testee, described mainboard comprises pci interface chip 5, arm processor 6, RAM8, the level transferring chip 9 that CPLD consists of, described treating apparatus is numerically-controlled machine 7, the testee image passes through two cameras 1 by synchronous recording, gather by capture card 3, video frequency collection card is connected with system board by pci bus 4, and ARM chip 6 is by interrupt mode control pci interface chip 5, its exchanges data cushions by dual port RAM 8, described ARM8 obtains the laggard line algorithm of image information by pci interface 5 and processes, and the control routine of final generating three-dimensional lathe exports control routine to lathe 7 by USB interface.In processing procedure, the control routine of generation can be kept among the FLASH, also can be stored in the external storage by USB interface.Need a slice to be connected by the level transferring chip that CPLD consists of between ARM and RAM and the FLASH.
Embodiment 5:As a kind of improvement of the present invention, described projector adopts the LCD liquid crystal imaging, projector distance is set to 1.2 5.7 meters, liquid crystal board is of a size of 0.7 inch, the standard display resolution is 1024 * 768 pixels, nominal contrast 450: 1, nominal light 2000 lumens, camera lens F value 1.8 2.15,23 27.6 millimeters of focal lengths.
Embodiment 6:As a kind of improvement of the present invention, described camera adopts the CCD camera, adopt 1/3 inch digital imagery plate, it adopts 1/3 inch digital imagery plate, 440,000 pixels, the horizontal resolution of 535 line high-resolutions, minimal illumination 0.7 lumen, signal to noise ratio (S/N ratio) is 50dB, and built-in 10 bit digital processors (DSP) are supported automatic backlight compensation and Automatic white balance.All the other structures and advantage and embodiment 4 are identical.
The present invention can also be combined to form new embodiment with at least one and the embodiment 1 in embodiment 2, the 3 described technical characterictics, and the present invention can also be combined to form new embodiment with at least one and the embodiment 4 in embodiment 5, the 6 described technical characterictics.
Need to prove that above-described embodiment only is preferred embodiment of the present invention; be not used for limiting protection scope of the present invention; being equal to replacement or substituting of making on the basis of the above all falls into protection scope of the present invention, and protection scope of the present invention is as the criterion with claims.
Claims (7)
1. based on the reverse-engineering job operation of space encoding projection, it is characterized in that, said method comprising the steps of, 1) to being gathered object projecting space coding, obtain again the coordinate information of the space scale of testee by video means, 2) image that collects is processed, reconstruct also is optimized for surface model; 3) surface model that generates is carried out analytical calculation, generate cutter path.
2. the reverse-engineering job operation based on the space encoding projection according to claim 1 is characterized in that, in the described step 2 image is carried out pretreated process and comprises, the denoising of some cloud, point cloud compressing, hole repair and point cloud model distortion are optimized;
Described some cloud denoising mainly comprises rejects the singular point noise and utilizes the average drifting iteration to carry out the offset point adjustment, wherein reject singular point, in the noise spot data, some point is obviously and on every side to be in different range, in model, be presented as burr, directly reject for singular point; Described adjustment offset point departs from entopic noise spot for some, need to carry out the position adjustment, makes it more near the position of original curved surface; Described point cloud compressing is on the Voronoi figure of cloud data structure, utilizes fast marching algorithms constantly to expand sample area, utilizes simultaneously the solstics sampling condition to filter out qualified sampled point; Described hole is repaired and is mainly comprised, at first extract the hole edge feature point, utilize radial basis function to simulate curved surface on the hole according to the information of hole peripheral point subsequently, subsequently according to certain step-length, sample at curved surface, adopt Moving Least to adjust to resulting point, so that the curved surface after repairing can be connected with curved surface is smooth on every side; Described point cloud model distortion is optimized, that point cloud model is carried out Free Transform more to be adapted to the data model of processing, the characteristic of described Free Transform mainly is that the character by basis function shows, and introduces different basis functions, just can form the Free Transform of different qualities.
3. the reverse-engineering job operation based on the space encoding projection according to claim 1 is characterized in that, described step 3 comprises the cross section distance computation, and cutter location calculates, and the cutter spacing path generates, and numerical control code generates.
4. the reverse-engineering system of processing based on the space encoding projection is characterized in that, described system comprises harvester, treating apparatus, processing unit (plant); Harvester connects treating apparatus, and described treating apparatus connects processing unit (plant); Described harvester, treating apparatus, processing unit (plant) adopt serial mode to connect.
5. the reverse-engineering system of processing based on the space encoding projection according to claim 4, it is characterized in that, described harvester comprises camera, projector, capture card, described treating apparatus is embedded main board, described mainboard comprises pci interface chip, arm processor, RAM, the level transferring chip that CPLD consists of, described treating apparatus is numerically-controlled machine, the testee image passes through two cameras by synchronous recording, by the video frequency collection card collection, video frequency collection card is connected with system board by pci bus, the ARM chip is controlled pci interface chip by interrupt mode, and its exchanges data cushions by dual port RAM, and described ARM obtains the laggard line algorithm of image information by pci interface and processes, the control routine of final generating three-dimensional lathe exports control routine to processing unit (plant) by USB interface.
6. the reverse-engineering system of processing based on the space encoding projection according to claim 5, it is characterized in that, described projector adopts the LCD liquid crystal imaging, projector distance is set to 1.2 5.7 meters, and liquid crystal board is of a size of 0.7 inch, and the standard display resolution is 1024 * 768 pixels, nominal contrast 450: 1, nominal light 2000 lumens, camera lens F value 1.8 2.15,23 27.6 millimeters of focal lengths.
7. the reverse-engineering system of processing based on the space encoding projection according to claim 5, it is characterized in that, described camera adopts the CCD camera, adopts 1/3 inch digital imagery plate, and it adopts 1/3 inch digital imagery plate, 440,000 pixels, the horizontal resolution of 535 line high-resolutions, minimal illumination 0.7 lumen, signal to noise ratio (S/N ratio) is 50dB, built-in 10 bit digital processors are supported automatic backlight compensation and Automatic white balance.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100157427A CN103049622A (en) | 2013-01-16 | 2013-01-16 | Reverse engineering processing system and method based on space code projection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100157427A CN103049622A (en) | 2013-01-16 | 2013-01-16 | Reverse engineering processing system and method based on space code projection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103049622A true CN103049622A (en) | 2013-04-17 |
Family
ID=48062259
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013100157427A Pending CN103049622A (en) | 2013-01-16 | 2013-01-16 | Reverse engineering processing system and method based on space code projection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103049622A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104475971A (en) * | 2014-11-04 | 2015-04-01 | 龚传波 | Beam splitting laser multi-station time sharing cutting machine and machining method |
| CN107214591A (en) * | 2017-06-12 | 2017-09-29 | 中国航发哈尔滨东安发动机有限公司 | The automatic Polishing method of high-temperature alloy blades |
| CN108010002A (en) * | 2017-12-25 | 2018-05-08 | 浙江大学 | A kind of structuring point cloud denoising method based on adaptive implicit Moving Least Squares |
| CN109102569A (en) * | 2018-06-13 | 2018-12-28 | 东莞时谛智能科技有限公司 | A kind of reconstruct foot point cloud model processing method and system |
| CN109388100A (en) * | 2017-08-07 | 2019-02-26 | 株式会社松浦机械制作所 | The automatic production method of cutting path in the inner space of three-D moulding object |
| CN109522903A (en) * | 2018-08-31 | 2019-03-26 | 董箭 | TIN_DDM Feature line extracting method based on the transformation of bidirectional rolling ball |
| CN110415331A (en) * | 2019-06-14 | 2019-11-05 | 中国电建集团昆明勘测设计研究院有限公司 | Method for rapidly detecting, judging and repairing contour or hole based on point cloud data |
| CN111077844A (en) * | 2019-12-12 | 2020-04-28 | 南京航空航天大学 | Part accurate machining method based on measured data feature guidance |
| CN112183807A (en) * | 2019-07-03 | 2021-01-05 | 三星Sds株式会社 | Circular frame generation method and calculation device |
| CN116109788A (en) * | 2023-02-15 | 2023-05-12 | 张春阳 | Method for modeling and reconstructing solid piece |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN203038276U (en) * | 2013-01-16 | 2013-07-03 | 杨安康 | Reverse engineering processing system based on space encoding projection |
-
2013
- 2013-01-16 CN CN2013100157427A patent/CN103049622A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN203038276U (en) * | 2013-01-16 | 2013-07-03 | 杨安康 | Reverse engineering processing system based on space encoding projection |
Non-Patent Citations (3)
| Title |
|---|
| 姜宏: "大尺寸复杂零件反求关键技术研究及应用", 《中国博士学位论文全文数据库工程科技Ⅱ辑》, no. 11, 15 November 2012 (2012-11-15) * |
| 钟国宇: "面向三角网格模型的直接数控编程系统研究", 《中国优秀硕士学位论文全文数据库工程科技Ⅰ辑》, 15 September 2012 (2012-09-15), pages 4 * |
| 高福顺: "由点云生成网格曲面算法的研究", 《中国博士学位论文全文数据库基础科学辑》, 15 September 2010 (2010-09-15) * |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104475971B (en) * | 2014-11-04 | 2016-06-22 | 龚传波 | A kind of beam splitting laser multistation timesharing cutting machine and processing method |
| CN104475971A (en) * | 2014-11-04 | 2015-04-01 | 龚传波 | Beam splitting laser multi-station time sharing cutting machine and machining method |
| CN107214591A (en) * | 2017-06-12 | 2017-09-29 | 中国航发哈尔滨东安发动机有限公司 | The automatic Polishing method of high-temperature alloy blades |
| CN109388100A (en) * | 2017-08-07 | 2019-02-26 | 株式会社松浦机械制作所 | The automatic production method of cutting path in the inner space of three-D moulding object |
| CN108010002B (en) * | 2017-12-25 | 2021-06-01 | 浙江大学 | Structured point cloud denoising method based on adaptive implicit moving least square |
| CN108010002A (en) * | 2017-12-25 | 2018-05-08 | 浙江大学 | A kind of structuring point cloud denoising method based on adaptive implicit Moving Least Squares |
| CN109102569A (en) * | 2018-06-13 | 2018-12-28 | 东莞时谛智能科技有限公司 | A kind of reconstruct foot point cloud model processing method and system |
| CN109522903A (en) * | 2018-08-31 | 2019-03-26 | 董箭 | TIN_DDM Feature line extracting method based on the transformation of bidirectional rolling ball |
| CN109522903B (en) * | 2018-08-31 | 2022-10-25 | 中国人民解放军海军大连舰艇学院 | TIN _ DDM topographic feature line extraction method based on bidirectional rolling ball transformation |
| CN110415331B (en) * | 2019-06-14 | 2022-08-19 | 中国电建集团昆明勘测设计研究院有限公司 | Method for rapidly detecting, judging and repairing contour or hole based on point cloud data |
| CN110415331A (en) * | 2019-06-14 | 2019-11-05 | 中国电建集团昆明勘测设计研究院有限公司 | Method for rapidly detecting, judging and repairing contour or hole based on point cloud data |
| CN112183807A (en) * | 2019-07-03 | 2021-01-05 | 三星Sds株式会社 | Circular frame generation method and calculation device |
| CN112183807B (en) * | 2019-07-03 | 2024-04-16 | 三星Sds株式会社 | Circular frame generation method and computing device |
| CN111077844B (en) * | 2019-12-12 | 2020-09-22 | 南京航空航天大学 | Part accurate machining method based on measured data feature guidance |
| CN111077844A (en) * | 2019-12-12 | 2020-04-28 | 南京航空航天大学 | Part accurate machining method based on measured data feature guidance |
| CN116109788A (en) * | 2023-02-15 | 2023-05-12 | 张春阳 | Method for modeling and reconstructing solid piece |
| CN116109788B (en) * | 2023-02-15 | 2023-07-04 | 张春阳 | Method for modeling and reconstructing solid piece |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103049622A (en) | Reverse engineering processing system and method based on space code projection | |
| CN112927370B (en) | Three-dimensional building model construction method and device, electronic equipment and storage medium | |
| CN103759671B (en) | A kind of dental model three-dimensional surface data non-contact scanning method | |
| CN111035115B (en) | Sole gluing path planning method and device based on 3D vision | |
| CN106600680A (en) | Batch type refined three-dimensional modeling method of building frame model | |
| KR101425576B1 (en) | Method for acquiring and processing a three-dimensional data to product a precise wide-area scale model | |
| CN107146287B (en) | Two-dimensional projection image to threedimensional model mapping method | |
| CN110837715B (en) | Complex curved surface machining error compensation method based on reverse engineering technology | |
| CN104422396B (en) | Assembling product gap three dimensional analysis system and method | |
| CN106097433A (en) | Object industry and the stacking method of Image model and system | |
| CN104268940B (en) | MEMS structure reconstruction and detection method based on CT scanned images | |
| CN203038276U (en) | Reverse engineering processing system based on space encoding projection | |
| CN102722906B (en) | Feature-based top-down image modeling method | |
| CN117475110B (en) | Semantic three-dimensional reconstruction method and device for blade, electronic equipment and storage medium | |
| US8948498B1 (en) | Systems and methods to transform a colored point cloud to a 3D textured mesh | |
| Kumar et al. | Curve reconstruction of digitized surface using K-means algorithm | |
| CN102445147A (en) | Measuring head simulation system and method of image measurement machine | |
| CN115758496A (en) | Viewpoint planning method and system for three-dimensional measurement of blade | |
| CN113432558B (en) | Device and method for measuring irregular object surface area based on laser | |
| CN113806892A (en) | Reverse engineering-based propeller digital design and processing method | |
| Zheng et al. | Automatic 3D model reconstruction of the diamond object | |
| CN113129348A (en) | Monocular vision-based three-dimensional reconstruction method for vehicle target in road scene | |
| Yang et al. | An Improved SFS Method for Achieving Fast, High-Precision, and Widely Adaptable 3-D Reconstruction | |
| Guo et al. | Automatic and high-accuracy matching method for a blade inspection system integrating fringe projection profilometry and conoscopic holography | |
| CN109785441A (en) | A kind of intelligence O&M method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20130417 |
|
| RJ01 | Rejection of invention patent application after publication |