CN113554559B - Three-dimensional reconstruction method and device with multiple complex curved surface rotational symmetry models - Google Patents
Three-dimensional reconstruction method and device with multiple complex curved surface rotational symmetry models Download PDFInfo
- Publication number
- CN113554559B CN113554559B CN202110687886.1A CN202110687886A CN113554559B CN 113554559 B CN113554559 B CN 113554559B CN 202110687886 A CN202110687886 A CN 202110687886A CN 113554559 B CN113554559 B CN 113554559B
- Authority
- CN
- China
- Prior art keywords
- point cloud
- dimensional
- point
- arm
- points
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 47
- 230000007246 mechanism Effects 0.000 claims abstract description 51
- 230000005540 biological transmission Effects 0.000 claims abstract description 30
- 238000009616 inductively coupled plasma Methods 0.000 claims description 18
- 230000009466 transformation Effects 0.000 claims description 15
- 238000001914 filtration Methods 0.000 claims description 12
- 230000001276 controlling effect Effects 0.000 claims description 11
- 239000011159 matrix material Substances 0.000 claims description 11
- 239000013598 vector Substances 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 7
- 230000002146 bilateral effect Effects 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 6
- 230000009467 reduction Effects 0.000 claims description 6
- 238000013519 translation Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 230000001105 regulatory effect Effects 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 230000009471 action Effects 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 9
- 238000005259 measurement Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003638 chemical reducing agent Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
- B25J9/1602—Programme controls characterised by the control system, structure, architecture
- B25J9/1605—Simulation of manipulator lay-out, design, modelling of manipulator
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
- B25J9/1656—Programme controls characterised by programming, planning systems for manipulators
- B25J9/1664—Programme controls characterised by programming, planning systems for manipulators characterised by motion, path, trajectory planning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
- G06T7/344—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods involving models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
- G06T7/75—Determining position or orientation of objects or cameras using feature-based methods involving models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
- G06T2207/20028—Bilateral filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30164—Workpiece; Machine component
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30244—Camera pose
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mechanical Engineering (AREA)
- Robotics (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Automation & Control Theory (AREA)
- Quality & Reliability (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a three-dimensional reconstruction method and a device with a multiple complex curved surface rotational symmetry model, wherein the device comprises a modeling support platform, an image acquisition mechanism and a five-axis grabbing and lifting rotary mechanical arm, the modeling support platform is provided with a transmission mechanism, multiple complex curved surface rotationally symmetric pieces to be detected are arranged at the top of the transmission mechanism, the pieces to be detected are driven to rotate through the transmission mechanism, the image acquisition mechanism is arranged at the end part of the five-axis grabbing and lifting rotary mechanical arm and moves along with the five-axis grabbing and lifting rotary mechanical arm to scan the pieces to be detected at multiple angles; when the device is used, the to-be-detected piece is driven to rotate through the modeling support platform, the to-be-detected piece can be scanned in a multi-angle and all-dimensional mode, two obtained point cloud images which are overlapped in pairs are roughly registered, then the point cloud images are precisely registered, and the spatial position difference between the point clouds is minimized, so that a three-dimensional model with a multiple complex curved surface rotation symmetry model is obtained, and the device has the characteristics of all-dimensional scanning, high scanning precision and convenience in use.
Description
Technical Field
The invention relates to the technical field of three-dimensional scanning reconstruction models, in particular to a three-dimensional reconstruction method and a three-dimensional reconstruction device with a multiple complex curved surface rotational symmetry model.
Background
The three-dimensional scanning technology is to acquire three-dimensional coordinate information of the surface of an object by utilizing certain physical phenomena such as light, sound, electromagnetism and the like which are bionically interacted with the surface of the object; with the continuous development of scientific technology, the three-dimensional scanning technology has been widely applied in the fields of cultural relics restoration, industrial engineering, natural disaster investigation and the like due to the advantages of high measurement speed, high precision, non-contact, convenient use and the like;
because the rotationally symmetrical model with multiple complex curved surfaces has multiple complex curved surfaces and the model is rotationally symmetrical, the model has the characteristic of overlapping shielding during scanning, complete three-dimensional surface information cannot be obtained through one-time scanning, and multiple multi-angle measurements are needed; in the scanning technology in the market at present, the obtained device with the multiple complex curved surface rotational symmetry model has the defects of complex operation, incapability of accurately controlling the scanning position of a three-dimensional scanner, overlong scanning time, easiness in missing scanning, excessive overlapping parts of point cloud depth images and the like.
Disclosure of Invention
Aiming at the existing problems, the invention aims to provide a three-dimensional reconstruction method and a device with a multiple complex curved surface rotational symmetry model, when in use, the device can accurately control a three-dimensional scanner to scan a piece to be detected at multiple scanning positions, and a modeling support table drives the piece to be detected to rotate, so that the piece to be detected can be scanned in multiple angles and all directions, and the device has the characteristics of all-directional scanning, high scanning precision and convenience in use.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
the three-dimensional reconstruction method with the multiple complex curved surface rotational symmetry model comprises the following steps
S1, clamping a structured light three-dimensional scanner at the arm end of a five-axis grabbing and lifting rotary mechanical arm, and fixing a to-be-detected piece on a modeling support table;
s2, controlling a five-axis grabbing and lifting rotary mechanical arm through a motion planning controller, and enabling a structured light three-dimensional scanner at the arm end to scan and observe a plurality of positions of a piece to be detected;
s3, controlling a stepping motor on the modeling support table to act through a stepping motor controller, and enabling an output shaft with a support disc to rotate at a fixed angle and at a fixed time;
s4, controlling a five-axis grabbing and lifting rotary mechanical arm through a motion planning controller to drive the three-dimensional scanner to a next observation position, and repeating the step S3 until a sufficient point cloud depth image is obtained;
s5, preprocessing the depth point cloud image to obtain corresponding three-dimensional point cloud data;
s6, carrying out filtering and denoising treatment on the obtained three-dimensional point cloud data;
s7, carrying out coarse registration on the filtered and denoised point cloud data, calculating corresponding translation vectors and rotation matrixes, and registering the translation vectors and the rotation matrixes into a unified coordinate system;
and S8, inputting two pieces of point cloud data which are adjacent and partially overlapped after coarse registration, performing fine splicing on the point cloud data by using an ICP (inductively coupled plasma) algorithm, solving a corresponding optimal transformation relation, and eliminating redundant information to splice the three-dimensional point cloud data into a complete model.
Preferably, the preprocessing process in step S5 includes:
s501, setting the pixel coordinates of pixel points in the point cloud depth image as (u, v), and counting three-dimensional point cloudsAccording to the coordinate of (x) w ,y w ,z w ) The coordinates of the camera optics in the image coordinate system are (u) 0 ,v 0 ) The focal length of the camera in the x-axis and the y-axis is set as f x ,f y The distance between the measured target and the depth three-dimensional scanner is d 0 ；
S502, the conversion relation between the pixel coordinates of the point cloud depth image and the coordinate points of the point cloud three-dimensional data is as follows:
and S503, obtaining three-dimensional point cloud data of the point cloud depth image after the overlapping through the formula (1).
Preferably, the specific method for performing filtering and denoising processing on the three-dimensional point cloud data in step S6 includes:
s601, firstly, calculating each data point p in the point cloud depth image according to the three-dimensional point cloud data i K field point N k (p i )；
S602, then p is paired i Finding W from the proximity point c Parameter of (c) p j -p i I and W s Parameter |<n j ,n i >-1||，<n i ,p j -p i >；
Wherein: in the above formula, W c (x) And W s (y) is a Gaussian kernel function, and the formula is as follows:
s603, obtaining the data points after bilateral filtering as
Wherein: in the above formula, λ is a bilateral filtering factor, and the calculation formula is:
preferably, in the step S7, when the obtained multiple sets of pairwise overlapped three-dimensional point cloud data are subjected to coarse registration, the adopted method is a PFH algorithm based on local features, and the process of performing coarse registration on the three-dimensional point cloud data by using the PFH algorithm includes:
s701, selecting n feature points from a cloud P to be registered, wherein the distance between the feature points is larger than a preset minimum distance threshold value delta;
s702, one or more points with similar PFH characteristic sub of characteristic points are searched in the target point cloud Q and the point cloud P, and one characteristic point is randomly selected from the one or more points as a corresponding point of 2 pieces of point clouds;
and S703, calculating a rigid body transformation matrix between corresponding points, and then judging the performance of the current registration transformation by solving a distance error sum function (using a Huber function) after the corresponding points are transformed.
Preferably, the PFH feature extraction process in step S702 includes:
(1) Taking a certain characteristic point of the three-dimensional point cloud data as a center, acquiring x points (with radius r and r capable of being defined) in the neighborhood of the certain characteristic point, pairing every two x points to obtain x (x + 1)/2 point pairs;
(2) Establishing a corresponding local coordinate system for each group of point pairs:
(3) In order to obtain a variation relation p system of each group of point pairs, using the uvw coordinate system obtained in the step (2), p points of 2 point pairs are converted into p points s ,p t Normal n to s ,n t The angular relationship of (a) is expressed in the following 3 variables:
wherein: d is the Euclidean distance between two points in the point pair;
(4) And (4) combining the feature collection of x (x + 1)/2 point pairs in the feature point field, namely (alpha, phi, theta, d) obtained in the step (3) into a feature histogram to obtain PFH features.
Preferably, the ICP algorithm in step S8 is an improved ICP algorithm, which uses a normal feature space sampling method to reduce the number of matching point pairs, and the closest point pair matching strategy used in the method is a linearly optimized point-to-plane method, and the ICP algorithm includes the following steps:
s801, setting a loss function of an ICP algorithm as follows:
E opt ＝arg min E ∑((E·w i -b i )·n i ) 2 (4)；
wherein, in the above function formula, w is the source vertex,
b is the target vertex and b is the target vertex,
n is a normal vector of b; e and E opt A transformation matrix of all 4 x 4;
s802, calculating a transformation matrix E according to the formula (4), solving an objective function E by using a linear approximation method, namely when theta =0 degrees, cos theta is approximately equal to 1, sin theta is approximately equal to theta, and then alpha, beta, gamma is approximately equal to 0, so that a rotation matrix of the E can be obtained:
that is E can be represented as
The three-dimensional reconstruction method with the multiple complex curved surface rotational symmetry model is realized based on a three-dimensional reconstruction device with the multiple complex curved surface rotational symmetry model, the device comprises a modeling support table, an image acquisition mechanism and a five-axis grabbing and lifting rotary mechanical arm, the modeling support table is provided with a transmission mechanism, a piece to be detected with multiple complex curved surface rotational symmetry is mounted at the top of the transmission mechanism, the piece to be detected is driven to rotate through the transmission mechanism, the image acquisition mechanism is arranged at the end part of the five-axis grabbing and lifting rotary mechanical arm and moves along with the five-axis grabbing and lifting rotary mechanical arm, and multi-angle scanning is carried out on the piece to be detected; the five-axis grabbing and lifting rotary mechanical arm comprises an arm seat, a first rotary arm, a second rotary arm, a rotary arm and a mechanical gripper, one end of the first rotary arm is connected with the arm seat through a first driving air pipe, the other end of the first rotary arm is connected with the second rotary arm through a second driving air pipe, the second rotary arm is connected with the rotary arm through a rotary connecting piece, and the mechanical gripper is arranged at the tail end of the rotary arm.
Preferably, the mechanical grippers are symmetrically arranged, the two mechanical grippers are connected through a connecting plate, and the connecting plate is connected with the rotating arm through a rotating rod; the mechanical gripper comprises symmetrically arranged side clamping pieces, a connecting block and a supporting plate, wherein the side clamping pieces are symmetrically arranged on two sides of the connecting block and are connected with the connecting block through a guide shaft and a telescopic shaft; the side clamping pieces are symmetrically provided with clamping blocks, the clamping blocks are also provided with a plurality of clamping grooves, and the clamping grooves are matched with the image acquisition mechanism for use; the side clamping piece at one side of the connecting block is fixedly connected with the connecting block through a connecting hinge plate, the side clamping piece at the other side of the connecting block is movably connected with the connecting block, a telescopic shaft at the end is a piston of a hydraulic cylinder, and the hydraulic cylinder is arranged in the connecting block; the connecting block passes through the connecting rod and is connected with the connecting plate, the lower extreme of connecting rod still is provided with the regulating block, the regulating block is in the dorsal spout of connecting block, and slides along the spout.
Preferably, the modeling support table further comprises a support mechanism, the support mechanism comprises support columns and support plates, the support columns are symmetrically arranged on the lower side of the support plate, the support plate is connected with the transmission mechanism and rotates along with the transmission mechanism, a connecting hole is formed in the support plate and is connected with a rotary output shaft, a connecting disc is further arranged at the top of the rotary output shaft, and the to-be-tested piece is connected with the rotary output shaft through the connecting disc; drive mechanism sets up the downside in the backup pad, including step motor, reduction gear and transmission shaft, step motor is controlled by step motor controller, and step motor passes through the reduction gear and is connected with the transmission shaft, the upper end and the backup pad of transmission shaft are connected, drive the backup pad through step motor and rotate.
Preferably, the image acquisition mechanism is a surface structured light three-dimensional scanner.
The invention has the beneficial effects that: the invention discloses a three-dimensional reconstruction method and a device with a multiple complex curved surface rotational symmetry model, compared with the prior art, the improvement of the invention is as follows:
the invention designs the three-dimensional reconstruction method and the device with the multiple complex curved surface rotational symmetry model, the device can accurately control the three-dimensional scanner to scan a to-be-detected piece at multiple scanning positions, the to-be-detected piece is driven to rotate by the modeling support platform, so that the to-be-detected piece is scanned in multiple angles and all around, a large number of point cloud depth images are obtained, and finally, a complete three-dimensional model with the multiple complex curved surface rotational symmetry model is obtained by point cloud splicing.
Drawings
FIG. 1 is a flow chart of a three-dimensional reconstruction method with a multiple complex surface rotationally symmetric model according to the present invention.
Fig. 2 is a schematic structural diagram I of the modeling support table of the present invention.
Fig. 3 is a schematic structural diagram II of the modeling support table of the present invention.
Fig. 4 is a schematic structural diagram of a stepping motor controller according to the present invention.
FIG. 5 is a schematic structural diagram of a five-axis gripper rotary robot arm according to the present invention.
FIG. 6 is a schematic structural diagram of the front end part of a five-axis grabbing and lifting rotating mechanical arm.
Fig. 7 is a structural schematic diagram of a front view angle of the mechanical gripper of the present invention.
Fig. 8 is a schematic view of a rear view of the mechanical gripper of the present invention.
Fig. 9 is a schematic structural diagram of the surface structured light three-dimensional scanner according to the present invention.
FIG. 10 is a diagram I of the effect of the three-dimensional reconstruction method with multiple rotationally symmetric complex surfaces according to example 1 of the present invention.
FIG. 11 is a diagram II of the effect of the model obtained by the three-dimensional reconstruction method with the multiple complex surface rotationally symmetric model in example 1 of the present invention.
Wherein: 1. the device comprises a to-be-detected piece, 2, a connecting disc, 3, a rotary output shaft, 4, a supporting column, 5, a stepping motor, 6, a supporting plate, 61, a connecting hole, 7, a speed reducer, 8, a transmission shaft, 9, a stepping motor controller, 10, an arm seat, 11, a first rotating arm, 12, a second rotating arm, 13, a first driving air pipe, 14, a second driving air pipe, 15, a rotating connecting piece, 16, a mechanical hand grip, 161, a connecting plate, 162, a side clamping piece, 163, a supporting plate, 164, a guide shaft, 165, a clamping block, 1651, a clamping groove, 166, a connecting block, 1661, a sliding groove, 167, a connecting rod, 1671, an adjusting block, 168, a telescopic shaft, 169, a connecting hinged plate, 17, a rotating rod, 18, a surface structure light three-dimensional scanner and 19.
Detailed Description
In order to make those skilled in the art better understand the technical solution of the present invention, the following description will be made with reference to the accompanying drawings and embodiments.
Referring to the attached figures 1-9, the three-dimensional reconstruction device with the multiple complex curved surface rotational symmetry model comprises a modeling support platform, an image acquisition mechanism and a five-axis grabbing and lifting rotary mechanical arm, wherein a transmission mechanism is arranged on the modeling support platform, a piece to be detected 1 with multiple complex curved surface rotational symmetry is installed at the top of the transmission mechanism, the transmission mechanism drives the piece to be detected 1 to rotate, the image acquisition mechanism is arranged at the end part of the five-axis grabbing and lifting rotary mechanical arm and moves along with the five-axis grabbing and lifting rotary mechanical arm to perform multi-angle scanning on the piece to be detected 1, so that multi-angle and all-dimensional scanning on the piece to be detected 1 is realized, a large amount of point cloud depth images are obtained, and later-stage model reconstruction and processing are facilitated.
Preferably, the five-axis grabbing and lifting rotary mechanical arm comprises an arm base 10, a first rotary arm 11, a second rotary arm 12, a rotary arm 19 and a mechanical gripper 16, one end of the first rotary arm 11 is connected with the arm base 10 through a first driving air pipe 13, the other end of the first rotary arm is connected with the second rotary arm 12 through a second driving air pipe 14, hydraulic pressure is transmitted through the first driving air pipe 13 and the second driving air pipe 14 to drive the first rotary arm 11 and the second rotary arm 12 to rotate, the second rotary arm 12 is connected with the rotary arm 19 through a rotary connecting piece 15, the mechanical gripper 16 is arranged at the tail end of the rotary arm 19, the rotary arm 19 is driven to rotate through external driving, and the mechanical gripper 16 is matched with an image acquisition mechanism for use.
Preferably, in order to facilitate clamping of the image capturing mechanism, the mechanical grippers 16 are symmetrically arranged, and the two mechanical grippers 16 are connected by a connecting plate 161, that is, the mechanical grippers 16 are symmetrically arranged at two ends of the connecting plate 161, and the connecting plate 161 is connected with the rotating arm 19 by a rotating rod 17.
Preferably, the mechanical gripper 16 includes symmetrically disposed side clamping pieces 162, a connecting block 166 and a supporting plate 163, the side clamping pieces 162 are symmetrically disposed on two sides of the connecting block 166, and are connected to the connecting block 166 through a guiding shaft 164 and a telescopic shaft 168, the supporting plate 163 is disposed on the lower side of the connecting block 166, the supporting plate 163 and the side clamping pieces 162 on two sides form a clamping groove, and the clamping groove is used in cooperation with the image capturing mechanism to clamp the image capturing mechanism.
Preferably, the side clamping pieces 162 are symmetrically provided with clamping blocks 165, and the clamping blocks 165 are further provided with a plurality of clamping grooves 1651, so that when the side clamping pieces are used, the clamping blocks 165 and the clamping grooves 1651 are matched with an image capturing mechanism for use to clamp the image capturing mechanism.
Preferably, the side clamping piece 162 on one side of the connecting block 166 is fixedly connected with the connecting block 166 through a connecting hinge plate 169, the side clamping piece 162 on the other side of the connecting block 166 is movably connected with the connecting block 166, the telescopic shaft 168 at the end is a piston of a hydraulic cylinder, and the hydraulic cylinder is arranged in the connecting block 166, namely when in use, the telescopic shaft 168 is driven by the hydraulic cylinder to move, so that the side clamping piece 162 is driven to move, the space of the clamping groove is enlarged or reduced, and the image acquisition mechanism is clamped and taken down conveniently.
Preferably, the connecting block 166 is connected with the connecting plate 161 through a connecting rod 167, the lower end of the connecting rod 167 is further provided with an adjusting block 1671, and the adjusting block 1671 is clamped in a sliding groove 1661 on the back side of the connecting block 166 and slides along the sliding groove 1661 to adjust the left and right spacing. To facilitate the connection of the junction block 166 with the junction plate 161.
Preferably, the image acquisition mechanism is a surface structured light three-dimensional scanner 18.
Preferably, the modeling support table further comprises a supporting mechanism, the supporting mechanism comprises a supporting column 4 and a supporting plate 6, the supporting column 4 is symmetrically arranged on the lower side of the supporting plate 6 to form a stable supporting leg, the supporting plate 6 is connected with a transmission mechanism, rotates along with the transmission mechanism under the drive of the transmission mechanism, a connecting hole 61 is formed in the supporting plate 6, the connecting hole 61 is connected with a rotary output shaft 3, the top of the rotary output shaft 3 is further provided with a connecting disc 2, a part to be tested 1 and the rotary output shaft 3 are connected through a flange of the connecting disc 2, namely, the part to be tested 1 is driven by the driving mechanism to rotate, and 360-degree dead angle-free scanning of the part to be tested 1 is achieved.
Preferably, drive mechanism set up the downside at backup pad 6, including step motor 5, reduction gear 7 and transmission shaft 8, step motor 5 is controlled by step motor controller 9, and step motor 5 passes through reduction gear 7 and is connected with transmission shaft 8, the upper end and the backup pad 6 of transmission shaft 8 are connected, drive backup pad 6 through step motor 5 and rotate.
The use process of the three-dimensional reconstruction device with the multiple complex curved surface rotational symmetry model comprises the following steps: (1) According to actual measurement requirements, firstly placing a five-axis grabbing and lifting rotary mechanical arm beside a modeling support table, and controlling the five-axis grabbing and lifting rotary mechanical arm to drive a surface structure light three-dimensional scanner by using a motion planning controller, so that the surface structure light three-dimensional scanner on the five-axis grabbing and lifting rotary mechanical arm is located at an optimal scanning position; (2) When the rotationally symmetric model with the multiple complex curved surfaces is scanned and modeled, the rotationally symmetric model with the multiple complex curved surfaces, which is fixedly connected to a supporting disc of a modeling supporting table, rotates by the same angle at intervals of the same time along with a preset program of a stepping motor controller on the modeling supporting table; (3) Scanning a model by a three-dimensional scanner at the arm end of the five-axis grabbing and lifting rotary mechanical arm according to the rotation frequency of a rotary supporting plate on the modeling supporting table through a preset program; (4) And completing scanning until the rotationally symmetrical model with the multiple complex curved surfaces rotates for a circle.
Note that: the three-dimensional scanner at the arm end of the five-axis grabbing and lifting rotary mechanical arm cannot completely scan the rotationally symmetric model with multiple complex curved surfaces at the same position, so that a complete reconstruction model cannot be obtained; at the moment, the five-axis grabbing and lifting rotary mechanical arm is controlled by the motion planning controller to drive the surface structured light three-dimensional scanner to the next observation position, and the scanned model is continuously scanned until a sufficient point cloud depth image is obtained.
The three-dimensional reconstruction method with the multiple complex curved surface rotational symmetry model by utilizing the three-dimensional reconstruction device with the multiple complex curved surface rotational symmetry model comprises the following steps
S1, clamping a structured light three-dimensional scanner at the arm end of a five-axis grabbing and lifting rotary mechanical arm, and fixing a to-be-detected piece on a modeling support table;
s2, controlling a five-axis grabbing and lifting rotary mechanical arm through a motion planning controller, and enabling a structured light three-dimensional scanner at the arm end to scan and observe a plurality of positions of a piece to be detected;
s3, controlling the action of a stepping motor on the modeling support table through a stepping motor controller to enable an output shaft with a support disc to rotate at a fixed angle and at a fixed time;
s4, controlling a five-axis grabbing and lifting rotary mechanical arm through a motion planning controller to drive the three-dimensional scanner to a next observation position, and repeating the step S3 until a sufficient point cloud depth image is obtained;
s5, preprocessing the depth point cloud image to obtain corresponding three-dimensional point cloud data;
s6, carrying out filtering and denoising treatment on the obtained three-dimensional point cloud data;
s7, carrying out coarse registration on the filtered and denoised point cloud data, calculating corresponding translation vectors and rotation matrixes, and registering the translation vectors and the rotation matrixes into a unified coordinate system;
and S8, inputting two pieces of point cloud data which are adjacent and partially overlapped after coarse registration, performing fine splicing on the point cloud data by using an ICP (inductively coupled plasma) algorithm, solving a corresponding optimal transformation relation, and eliminating redundant information to splice the three-dimensional point cloud data into a complete model.
Preferably, the preprocessing process in step S5 includes:
s501, setting the pixel coordinates of pixel points in the cloud depth image as (u, v), and setting the coordinates of three-dimensional point cloud data as (x) w ,y w ,z w ) The coordinates of the camera optics in the image coordinate system are (u) 0 ,v 0 ) The focal length of the camera in the x-axis and the y-axis is set as f x ,f y The distance between the measured target and the depth three-dimensional scanner is d 0 ；
S502, the conversion relation between the pixel coordinates of the point cloud depth image and the coordinate points of the point cloud three-dimensional data is as follows:
and S503, obtaining three-dimensional point cloud data of the point cloud depth image after the overlapping through the formula (1).
Preferably, the specific method for performing filtering and denoising processing on the three-dimensional point cloud data in step S6 includes:
s601, firstly, calculating each data point p in the point cloud depth image according to the three-dimensional point cloud data i K field point N k (p i )；
S602, then p is paired i Obtaining W from the proximity point of c Parameter of (1 | | p) j -p i I and W s Parameter |<n j ,n i >-1||，<n i ,p j -p i >；
Wherein: in the above formula, W c (x) And W s (y) is a Gaussian kernel function, and the formula is as follows:
s603, obtaining the data points after bilateral filtering
Wherein: in the above formula, λ is a bilateral filtering factor, and the calculation formula is:
preferably, in the step S7, when the obtained multiple sets of pairwise overlapped three-dimensional point cloud data are subjected to coarse registration, the adopted method is a PFH algorithm based on local features, and the process of performing coarse registration on the three-dimensional point cloud data by using the PFH algorithm includes:
s701, selecting n feature points from a cloud P to be registered, wherein the distance between the feature points is larger than a preset minimum distance threshold value delta;
s702, one or more points with similar PFH characteristic sub of the characteristic points are searched in the target point cloud Q and the point cloud P, one characteristic point is randomly selected from the one or more points as the corresponding point of 2 pieces of point clouds, and the extraction process of the PFH characteristic sub comprises the following steps:
(1) Taking a certain characteristic point of the three-dimensional point cloud data as a center, acquiring x points (with radius r and r capable of being defined) in the neighborhood of the certain characteristic point, and pairing the x points pairwise to obtain x (x + 1)/2 point pairs;
(2) Establishing a corresponding local coordinate system for each group of point pairs:
(3) In order to obtain a variation relation p system of each group of point pairs, 2 points p in the point pairs are utilized by the uvw coordinate system obtained in the step (2) s ,p t Normal n to s ,n t Is expressed in the following 3 variables:
wherein: d is the Euclidean distance between two points in the point pair;
(4) Combining the feature collection of x (x + 1)/2 point pairs in the feature point field, namely (alpha, phi, theta, d) obtained in the step (3) into a feature histogram to obtain PFH features;
and S703, calculating a rigid body transformation matrix between corresponding points, and then judging the performance of the current registration transformation by solving a distance error sum function (using a Huber function) after the corresponding points are transformed.
Preferably, the ICP algorithm in step S8 is an improved ICP algorithm, which uses a normal feature space sampling method to reduce the number of matching point pairs, and the closest point pair matching strategy used in the method is a linearly optimized point-to-plane method, and the ICP algorithm includes the following steps:
s801, setting a loss function of an ICP algorithm as follows:
E opt ＝arg min E ∑((E·w i -b i )·n i ) 2 (4)；
wherein, in the above function formula, w is the source vertex,
b is the target vertex and b is the target vertex,
n is a normal vector of b; e and E opt A transformation matrix of all 4 x 4;
s802, calculating a transformation matrix E according to the formula (4), solving an objective function E by using a linear approximation method, namely when theta =0 degrees, cos theta is approximately equal to 1, sin theta is approximately equal to theta, and alpha, beta, gamma is approximately equal to 0, so that a rotation matrix of the E can be obtained:
that is E can be represented as
Example 1:
placing a five-axis grabbing and lifting rotary mechanical arm beside a modeling support table, opening a power supply device of the five-axis grabbing and lifting rotary mechanical arm, utilizing a motion planning controller to clamp a surface structure light three-dimensional scanner on a mechanical gripper at the arm end of the mechanical arm, planning a motion track of the five-axis grabbing and lifting rotary mechanical arm through the motion planning controller according to task requirements, and enabling the surface structure light three-dimensional scanner on the five-axis grabbing and lifting rotary mechanical arm to move to multiple optimal scanning positions;
fixedly connecting a piece to be measured on a rotating output shaft of a modeling support table, inputting a preset program into a stepping motor controller, starting a stepping motor on the modeling support table, and enabling the rotating output shaft on the modeling support table to rotate for the same angle at intervals of the same time until the rotating output shaft rotates for a circle, and stopping rotating;
the method comprises the following steps that a three-dimensional scanner at the arm end of a five-axis grabbing and lifting rotary mechanical arm scans a model once every time the model rotates according to the rotating frequency of a rotary output shaft on a modeling support table through a preset program until a piece to be detected rotates for one circle, and scanning is finished;
after the first scanning is finished, the five-axis grabbing and lifting rotary mechanical arm drives the surface structured light three-dimensional scanner to move to the next scanning position, the transmission mechanism of the modeling support table is restarted, and the previous step is repeated.
And scanning for multiple times to obtain a complete point cloud depth image, uploading the complete point cloud depth image to an upper computer, and performing operation processing on the obtained depth image through a preset algorithm program to obtain a complete three-dimensional model of the to-be-detected piece, as shown in fig. 10 and 11.
The foregoing shows and describes the general principles, principal features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (8)
1. The three-dimensional reconstruction method with the multiple complex curved surface rotational symmetry model is characterized by comprising the following steps of: comprises the steps of
S1, clamping a structured light three-dimensional scanner at the arm end of a five-axis grabbing and lifting rotary mechanical arm, and fixing a to-be-detected piece on a modeling support table;
s2, controlling a five-axis grabbing and lifting rotary mechanical arm through a motion planning controller, and enabling a structured light three-dimensional scanner at the arm end to scan and observe a plurality of positions of a piece to be detected;
s3, controlling the action of a stepping motor on the modeling support table through a stepping motor controller to enable an output shaft with a support disc to rotate at a fixed angle and at a fixed time;
s4, controlling a five-axis grabbing and lifting rotary mechanical arm through a motion planning controller to drive the three-dimensional scanner to a next observation position, and repeating the step S3 until a sufficient point cloud depth image is obtained;
s5, preprocessing the depth point cloud image to obtain corresponding three-dimensional point cloud data;
s6, carrying out filtering and denoising treatment on the obtained three-dimensional point cloud data;
s7, carrying out coarse registration on the filtered and denoised point cloud data, calculating corresponding translation vectors and rotation matrixes, and registering the corresponding translation vectors and the rotation matrixes into a unified coordinate system;
s8, inputting two pieces of point cloud data which are adjacent and partially overlapped after coarse registration, performing fine splicing on the point cloud data by using an ICP (inductively coupled plasma) algorithm, solving a corresponding optimal transformation relation, and eliminating redundant information at the same time to splice the three-dimensional point cloud data into a complete model;
the method is realized based on a three-dimensional reconstruction device with a multiple complex curved surface rotational symmetry model, the device comprises a modeling support table, an image acquisition mechanism and a five-axis grabbing and lifting rotary mechanical arm, a transmission mechanism is arranged on the modeling support table, a multiple complex curved surface rotationally symmetric to-be-detected piece (1) is installed at the top of the transmission mechanism, the to-be-detected piece (1) is driven to rotate through the transmission mechanism, the image acquisition mechanism is arranged at the end part of the five-axis grabbing and lifting rotary mechanical arm and moves along with the five-axis grabbing and lifting rotary mechanical arm, and multi-angle scanning is carried out on the to-be-detected piece (1); the five-axis grabbing and lifting rotary mechanical arm comprises an arm seat (10), a first rotary arm (11), a second rotary arm (12), a rotary arm (19) and a mechanical gripper (16), wherein one end of the first rotary arm (11) is connected with the arm seat (10) through a first driving air pipe (13), the other end of the first rotary arm is connected with the second rotary arm (12) through a second driving air pipe (14), the second rotary arm (12) is connected with the rotary arm (19) through a rotary connecting piece (15), and the mechanical gripper (16) is arranged at the tail end of the rotary arm (19);
the mechanical grippers (16) are symmetrically arranged, the two mechanical grippers (16) are connected through a connecting plate (161), and the connecting plate (161) is connected with the rotating arm (19) through a rotating rod (17); the mechanical gripper (16) comprises symmetrically arranged side clamping pieces (162), a connecting block (166) and a supporting plate (163), wherein the side clamping pieces (162) are symmetrically arranged on two sides of the connecting block (166) and are connected with the connecting block (166) through a guide shaft (164) and a telescopic shaft (168), the supporting plate (163) is arranged on the lower side of the connecting block (166), the supporting plate (163) and the side clamping pieces (162) on two sides form a clamping groove, and the clamping groove is matched with the image acquisition mechanism for use; clamping blocks (165) are symmetrically arranged on the side clamping piece (162), a plurality of clamping grooves (1651) are further formed in the clamping blocks (165), and the clamping grooves (1651) are matched with the image acquisition mechanism for use; the side clamping piece (162) on one side of the connecting block (166) is fixedly connected with the connecting block (166) through a connecting hinged plate (169), the side clamping piece (162) on the other side of the connecting block (166) is movably connected with the connecting block (166), a telescopic shaft (168) at the end is a piston of a hydraulic cylinder, and the hydraulic cylinder is arranged in the connecting block (166); connecting block (166) are connected with connecting plate (161) through connecting rod (167), the lower extreme of connecting rod (167) still is provided with regulating block (1671), regulating block (1671) block is in connecting block (166) dorsal spout (1661), and slides along spout (1661).
2. The three-dimensional reconstruction method with multiple complex curved surface rotational symmetry models according to claim 1, characterized in that: the preprocessing process of step S5 includes:
s501, setting the pixel coordinates of pixel points in the cloud depth image as (u, v), and setting the coordinates of the three-dimensional point cloud data as (x) w ,y w ,z w ) The coordinates of the camera optics in the image coordinate system are (u) 0 ,v 0 ) The focal length of the camera in the x-axis and the y-axis is set as f x ,f y The distance between the measured target and the depth three-dimensional scanner is d 0 ；
S502, the conversion relation between the pixel coordinates of the point cloud depth image and the coordinate points of the point cloud three-dimensional data is as follows:
s503, obtaining three-dimensional point cloud data of the point cloud depth image after the overlapping is taken out through the formula (1).
3. The three-dimensional reconstruction method with multiple complex curved surface rotational symmetry models according to claim 1, characterized in that: the specific method for performing filtering and denoising processing on the three-dimensional point cloud data in the step S6 includes:
s601, firstly, calculating each data point p in the point cloud depth image according to the three-dimensional point cloud data i K neighborhood point N k (p i )；
S602, then p is paired i Finding W of adjacent points c Parameter of (1 | | p) j -p i I and W s Parameter |<n j ,n i >-1||，<n i ,p j -p i >；
Wherein: in the above formula, W c (x) And W s (y) is a Gaussian kernel function, and the formula is as follows:
s603, obtaining the data points after bilateral filtering
Wherein: in the above formula, λ is a bilateral filtering factor, and the calculation formula is:
4. the three-dimensional reconstruction method with multiple complex curved surface rotational symmetry models according to claim 1, characterized in that: in the step S7, when the obtained multiple sets of pairwise overlapped three-dimensional point cloud data are coarsely registered, the adopted method is a PFH algorithm based on local features, and the process of coarsely registering the three-dimensional point cloud data by using the PFH algorithm includes:
s701, selecting n feature points from a cloud P to be registered, wherein the distance between the feature points is larger than a preset minimum distance threshold value delta;
s702, one or more points with similar PFH characteristic sub of characteristic points are searched in the target point cloud Q and the point cloud P, and one characteristic point is randomly selected from the one or more points as a corresponding point of 2 pieces of point clouds;
and S703, calculating a rigid transformation matrix between the corresponding points, and then judging the performance of the current registration transformation by solving the distance error and the function after the corresponding points are transformed.
5. The three-dimensional reconstruction method with multiple complex curved surface rotational symmetry models according to claim 4, characterized in that: step S702, the extraction process of the PFH feature includes:
(1) Taking a certain characteristic point of the three-dimensional point cloud data as a center, acquiring x points in the neighborhood of the certain characteristic point, pairing the x points pairwise to obtain x (x + 1)/2 point pairs;
(2) Establishing a corresponding local coordinate system for each group of point pairs:
(3) In order to obtain the variation relation of each group of point pairs, 2 points p in the point pairs are utilized by the uvw coordinate system obtained in the step (2) s ,p t Normal n of s ,n t The angular relationship of (a) is expressed in the following 3 variables:
wherein: d is the Euclidean distance between two points in the point pair;
(4) And (3) merging the feature collections of x (x + 1)/2 point pairs in the feature point field, namely (alpha, phi, theta, d) obtained in the step (3) into a feature histogram to obtain PFH features.
6. The three-dimensional reconstruction method with multiple complex surface rotationally symmetric models according to claim 1, characterized in that: the ICP algorithm in step S8 is an improved ICP algorithm, which uses a normal feature space sampling method to reduce the number of matching point pairs, and the closest point pair matching strategy used is a linearly optimized point-to-plane method, and the ICP algorithm includes the following steps:
s801, setting a loss function of an ICP algorithm as follows:
E opt ＝argmin E ∑((E·w i -b i )·n i ) 2 (4)；
wherein, in the above function formula, w is the source vertex,
b is the target vertex and b is the target vertex,
n is a normal vector of b; e and E opt 4 x 4 transformation matrices;
s802, calculating a transformation matrix E according to the formula (4), solving an objective function E by using a linear approximation method, namely when theta =0 degrees, cos theta is approximately equal to 1, sin theta is approximately equal to theta, and then alpha, beta, gamma is approximately equal to 0, so that a rotation matrix of the E can be obtained:
that is E can be represented as
7. The three-dimensional reconstruction method with multiple complex curved surface rotational symmetry models according to claim 1, characterized in that: the modeling support table further comprises a support mechanism, the support mechanism comprises support columns (4) and support plates (6), the support columns (4) are symmetrically arranged on the lower sides of the support plates (6), the support plates (6) are connected with a transmission mechanism and rotate along with the transmission mechanism, connecting holes (61) are formed in the support plates (6), the connecting holes (61) are connected with a rotary output shaft (3), a connecting disc (2) is further arranged at the top of the rotary output shaft (3), and a to-be-tested piece (1) is connected with the rotary output shaft (3) through the connecting disc (2); drive mechanism sets up the downside in backup pad (6), including step motor (5), reduction gear (7) and transmission shaft (8), step motor (5) are controlled by step motor controller (9), and step motor (5) are connected with transmission shaft (8) through reduction gear (7), the upper end and the backup pad (6) of transmission shaft (8) are connected, drive backup pad (6) through step motor (5) and rotate.
8. The three-dimensional reconstruction method with multiple complex curved surface rotational symmetry models according to claim 7, characterized in that: the image acquisition mechanism is a surface structured light three-dimensional scanner (18).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110687886.1A CN113554559B (en) | 2021-06-21 | 2021-06-21 | Three-dimensional reconstruction method and device with multiple complex curved surface rotational symmetry models |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110687886.1A CN113554559B (en) | 2021-06-21 | 2021-06-21 | Three-dimensional reconstruction method and device with multiple complex curved surface rotational symmetry models |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113554559A CN113554559A (en) | 2021-10-26 |
| CN113554559B true CN113554559B (en) | 2023-03-10 |
Family
ID=78102227
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110687886.1A Active CN113554559B (en) | 2021-06-21 | 2021-06-21 | Three-dimensional reconstruction method and device with multiple complex curved surface rotational symmetry models |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113554559B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115690184B (en) * | 2022-10-24 | 2024-02-06 | 西南交通大学 | Tunnel face displacement measurement method based on three-dimensional laser scanning |
| CN117162098B (en) * | 2023-10-07 | 2024-05-03 | 合肥市普适数孪科技有限公司 | Autonomous planning system and method for robot gesture in narrow space |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108340398A (en) * | 2018-03-12 | 2018-07-31 | 天津朗硕机器人科技有限公司 | A kind of multi-functional handgrip of the robot 3C Product assembly of view-based access control model positioning |
| CN109272524A (en) * | 2018-08-27 | 2019-01-25 | 大连理工大学 | A kind of small scale point cloud noise denoising method based on Threshold segmentation |
| CN109916333A (en) * | 2019-04-04 | 2019-06-21 | 大连交通大学 | A kind of large scale target with high precision three-dimensional reconstruction system and method based on AGV |
| CN110070607A (en) * | 2019-04-02 | 2019-07-30 | 深圳市掌网科技股份有限公司 | Three-dimensional reconstruction system and method |
| CN110490912A (en) * | 2019-07-17 | 2019-11-22 | 哈尔滨工程大学 | 3D-RGB point cloud registration method based on local gray level sequence model descriptor |
| WO2020077660A1 (en) * | 2018-10-19 | 2020-04-23 | 广东飞码机器人科技有限公司 | Operating system and method for surface coating robot |
| CN111872927A (en) * | 2020-07-28 | 2020-11-03 | 北京轩宇智能科技有限公司 | Redundant mechanical arm and system for controlling redundant mechanical arm |
| CN112509063A (en) * | 2020-12-21 | 2021-03-16 | 中国矿业大学 | Mechanical arm grabbing system and method based on edge feature matching |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10278584B2 (en) * | 2013-03-11 | 2019-05-07 | Carestream Dental Technology Topco Limited | Method and system for three-dimensional imaging |
| US10360469B2 (en) * | 2015-01-15 | 2019-07-23 | Samsung Electronics Co., Ltd. | Registration method and apparatus for 3D image data |
| CN106959080B (en) * | 2017-04-10 | 2019-04-05 | 上海交通大学 | A kind of large complicated carved components three-dimensional pattern optical measuring system and method |
| CN109785440A (en) * | 2018-12-18 | 2019-05-21 | 合肥阿巴赛信息科技有限公司 | A kind of curved surface distorted pattern |
| CN112634435A (en) * | 2020-12-17 | 2021-04-09 | 中国地质大学（武汉） | Ceramic product three-dimensional point cloud reconstruction method based on Eye in Hand model |
-
2021
- 2021-06-21 CN CN202110687886.1A patent/CN113554559B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108340398A (en) * | 2018-03-12 | 2018-07-31 | 天津朗硕机器人科技有限公司 | A kind of multi-functional handgrip of the robot 3C Product assembly of view-based access control model positioning |
| CN109272524A (en) * | 2018-08-27 | 2019-01-25 | 大连理工大学 | A kind of small scale point cloud noise denoising method based on Threshold segmentation |
| WO2020077660A1 (en) * | 2018-10-19 | 2020-04-23 | 广东飞码机器人科技有限公司 | Operating system and method for surface coating robot |
| CN110070607A (en) * | 2019-04-02 | 2019-07-30 | 深圳市掌网科技股份有限公司 | Three-dimensional reconstruction system and method |
| CN109916333A (en) * | 2019-04-04 | 2019-06-21 | 大连交通大学 | A kind of large scale target with high precision three-dimensional reconstruction system and method based on AGV |
| CN110490912A (en) * | 2019-07-17 | 2019-11-22 | 哈尔滨工程大学 | 3D-RGB point cloud registration method based on local gray level sequence model descriptor |
| CN111872927A (en) * | 2020-07-28 | 2020-11-03 | 北京轩宇智能科技有限公司 | Redundant mechanical arm and system for controlling redundant mechanical arm |
| CN112509063A (en) * | 2020-12-21 | 2021-03-16 | 中国矿业大学 | Mechanical arm grabbing system and method based on edge feature matching |
Non-Patent Citations (4)
| Title |
|---|
| Iterative closest point registration for fast point feature histogram features of a volume density optimization algorithm;Lu-shenWu et al;《Measurement and Control》;20201231;第53卷;第29-39页 * |
| 保持特征的散乱点云数据去噪;崔鑫等;《光学精密工程》;20171231;第25卷(第12期);正文第2.4节 * |
| 基于RGB-D数据的点云配准;薛瑞;《中国优秀硕士学位论文全文数据库》;20180415;正文第4.1-4.2节 * |
| 面向复杂曲面工件的测量机械臂轨迹规划方法研究;何喆;《中国优秀硕士学位论文全文数据库信息科技辑》;20210115;正文第2.2节、第3.2节 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113554559A (en) | 2021-10-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111750805B (en) | Three-dimensional measuring device and method based on binocular camera imaging and structured light technology | |
| CN108921901B (en) | Large-view-field camera calibration method based on precise two-axis turntable and laser tracker | |
| CN107830813B (en) | The longaxones parts image mosaic and bending deformation detection method of laser wire tag | |
| CN113554559B (en) | Three-dimensional reconstruction method and device with multiple complex curved surface rotational symmetry models | |
| CN110238845B (en) | Automatic hand-eye calibration method and device for optimal calibration point selection and error self-measurement | |
| CN110815213B (en) | Part identification and assembly method and device based on multi-dimensional feature fusion | |
| WO2018103694A1 (en) | Robotic three-dimensional scanning device and method | |
| CN106017325B (en) | Improved non-contact optical measurement method for complex surface and irregular object volume | |
| CN105945909A (en) | Error correction method and system for three-degree-of-freedom parallel robot | |
| CN108534710A (en) | A kind of the three-D profile scanning means and method of single line laser | |
| CN112132908B (en) | Camera external parameter calibration method and device based on intelligent detection technology | |
| CN101458072A (en) | Three-dimensional contour outline measuring set based on multi sensors and measuring method thereof | |
| CN104807494A (en) | Object micro surface morphology optical five-degree-of-freedom measurement device and method | |
| CN111862238B (en) | Full-space monocular light pen type vision measurement method | |
| CN104079810A (en) | Three-dimensional panoramic scanning device and three-dimensional model generating method | |
| CN108253939A (en) | Variable optical axis single eye stereo vision measuring method | |
| CN114119456A (en) | Automatic pipeline centering method based on machine vision | |
| CN111951380B (en) | Three-dimensional reconstruction device and method based on monocular stereoscopic vision | |
| CN111833392A (en) | Multi-angle scanning method, system and device for mark points | |
| CN105004283A (en) | Location system for three-dimensional measurement of dental model and calibration method thereof | |
| TWI272370B (en) | System and method for image measuring | |
| CN116428982A (en) | Fiber ring skeleton non-contact measurement system and method based on line structured light | |
| CN114926539A (en) | Efficient three-dimensional visual positioning method and device for large casting polishing | |
| CN112361982B (en) | Method and system for extracting three-dimensional data of large-breadth workpiece | |
| CN115502940A (en) | Instrument suitable for three-dimensional detection of surface defects of precise optical element |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |