CN112561974A - PCL-based 360-degree three-dimensional human head point cloud registration method - Google Patents
PCL-based 360-degree three-dimensional human head point cloud registration method Download PDFInfo
- Publication number
- CN112561974A CN112561974A CN202011614875.2A CN202011614875A CN112561974A CN 112561974 A CN112561974 A CN 112561974A CN 202011614875 A CN202011614875 A CN 202011614875A CN 112561974 A CN112561974 A CN 112561974A
- Authority
- CN
- China
- Prior art keywords
- point cloud
- registration
- human head
- degree
- icp
- 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
- 238000000034 method Methods 0.000 title claims abstract description 19
- 238000009616 inductively coupled plasma Methods 0.000 claims abstract description 33
- 238000005070 sampling Methods 0.000 claims abstract description 17
- 230000000007 visual effect Effects 0.000 claims abstract description 11
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 230000009466 transformation Effects 0.000 claims description 8
- 238000012800 visualization Methods 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 5
- 238000012545 processing Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 2
- 238000013475 authorization Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 239000013078 crystal Substances 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000003860 storage Methods 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
- G06T7/30—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
-
- 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
- G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4038—Image mosaicing, e.g. composing plane images from plane sub-images
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a PCL-based 360-degree three-dimensional human head point cloud registration method. Firstly, reading PCD files of three different visual angles of a human head, wherein the PCD files are L, M, R files and correspond to the left side, the middle side and the right side of the human head; then, coarse registration based on sampling consistency is used for carrying out registration operation, and a better initial position is provided for ICP fine registration; and finally, carrying out accurate registration operation by using an ICP (inductively coupled plasma) algorithm. The 360-degree three-dimensional human head point cloud registration result can be accurately and rapidly obtained.
Description
Technical Field
The invention relates to the technical field of optical measurement, in particular to a PCL-based 360-degree three-dimensional human head point cloud registration method.
Background
PCL (PointCloudLibrary) is a large cross-platform open source C + + programming library established on the basis of absorbing the prior point cloud related research, realizes a large number of point cloud related general algorithms and efficient data structures, and relates to point cloud acquisition, filtering, segmentation, registration, retrieval, feature extraction, identification, tracking, curved surface reconstruction, visualization and the like. And the system supports various operating system platforms and can run on Windows, Linux, Android, MacOSX and partially embedded real-time systems. If OpenCV is a crystal for 2D information acquisition and processing, PCL is equivalent in 3D information acquisition and processing, and PCL is a BSD authorization method, and can be commercially and academic applied for free.
ICP point cloud registration is one of the common point cloud processing algorithms. In fact, point cloud data is widely used in shape detection and classification, stereo vision, motion recovery structures, and multi-view reconstruction. The storage, compression, rendering and other problems of the point cloud are also hot points of research. With the popularization of point cloud collection equipment, the development of binocular stereo vision technology, VR and AR, the point cloud data processing technology is becoming one of the most promising technologies. PCL is a necessary tool and basic skill in the field of three-dimensional point cloud data processing. Thus, PCL-based technologies are increasingly being deployed.
Disclosure of Invention
In order to obtain a 360-degree three-dimensional human head point cloud registration result, the invention provides a PCL-based 360-degree three-dimensional human head point cloud registration method.
The technical scheme of the invention is as follows: a PCL-based 360-degree three-dimensional human head point cloud registration method comprises the following steps:
reading PCD files of three different visual angles of a human head, wherein the PCD files are L, M, R files and correspond to the left side, the middle side and the right side of the human head;
performing registration operation by using coarse registration based on sampling consistency to provide a better initial position for ICP fine registration;
performing accurate registration operation by using an ICP (inductively coupled plasma) algorithm to obtain a 360-degree three-dimensional human head point cloud registration result;
and fourthly, visually displaying the registration result of the 360-degree three-dimensional human head point cloud.
Preferably, in the step one, the PCD files of the three different visual angles of the human head are read in, and the PCD files of the three different visual angles of the human head are loaded respectively by using io: (loadPCDFile < PointXYZ >).
Preferably, in step two, a consistency initial registration algorithm is sampled, and before the algorithm is executed, a fast point feature histogram FPFH of the point cloud is calculated, and the algorithm flow is as follows:
step 2.1, from point cloud to be registeredPSelecting n sampling points, wherein the distance between every two sampling points is greater than a preset minimum distance threshold value d in order to ensure that the sampled points have different FPFH characteristics as much as possible;
step 2.2. in the target point cloudQMiddle search and point cloudPOne or more points with similar FPFH characteristics are sampled at the middle sampling point, and one point is randomly selected from the similar points to be used as a point cloudPIn the target point cloudQOne-to-one correspondence point of (1);
and 2.3, calculating a rigid body transformation matrix between the corresponding points, and then judging the performance of the current registration transformation result by solving the sum of the distance errors after the corresponding points are transformed.
Preferably, in step three, an ICP algorithm is used to perform a precise registration operation, where the ICP registration operation:
the ICP registration operation sets the following conditions: setting source point cloudsPWith a target point cloudQThe distance threshold of (2) is 7, and as long as the two points are smaller than the threshold, the two points are considered as corresponding points; setting a convergence judgment condition, a maximum allowable difference between two continuous conversions and a difference value between the last conversion and the current conversion; before the algorithm is considered to be converged, obtaining the difference value between the maximum allowable distance error and the error of two iterations; setting the maximum iteration number 30 of the optimized operation;
the ICP registration procedure is as follows: firstly, performing ICP registration of an M surface and an R surface; then, performing ICP registration of the M surface and the L surface; and finally, performing ICP registration of the R surface and the L surface to realize 360-degree three-dimensional human head point cloud registration.
Preferably, the step four is specifically: firstly, creating a point cloud visualization window; then, setting is carried out, wherein the right side is represented by green point cloud, the middle side is represented by red point cloud, and the left side is represented by white point cloud; and finally, splicing the point clouds which are registered pairwise together to realize 360-degree three-dimensional human head point cloud registration visualization.
Compared with the traditional method, the method has the following advantages: the invention discloses a PCL-based 360-degree three-dimensional human head point cloud registration method. Firstly, reading PCD files of three different visual angles of a human head, wherein the PCD files are L, M, R files and correspond to the left side, the middle side and the right side of the human head; then, coarse registration based on sampling consistency is used for carrying out registration operation, and a better initial position is provided for ICP fine registration; and finally, carrying out accurate registration operation by using an ICP (inductively coupled plasma) algorithm. The 360-degree three-dimensional human head point cloud registration result can be accurately and rapidly obtained.
Drawings
Fig. 1 is a schematic flow chart of a PCL-based 360-degree three-dimensional human head point cloud registration method in an embodiment of the present invention.
Fig. 2 is a flow chart of ICP registration in an embodiment of the invention.
Fig. 3 is a visual diagram of PCL-based 360 ° three-dimensional human head point cloud registration in an embodiment of the present invention. Wherein, the R surface is represented by green point cloud, the M surface is represented by red point cloud, and the L surface is represented by white point cloud.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
With reference to fig. 1 to 3, the embodiment provides a PCL-based 360-degree three-dimensional human head point cloud registration method, which includes the following five steps.
Reading PCD files of three different visual angles of a human head. Using io:: loadPCDFile < PointXYZ >, loading PCD files of three different visual angles of the human head respectively. And if the loading is successful, performing the operation of the second step.
And step two, coarse registration based on sampling consistency is used.
Sampling consistency initial registration algorithm. The algorithm relies on the histogram of point features, so before executing the algorithm, the fast histogram of point features FPFH of the point cloud should be calculated, and the algorithm flow is as follows:
(1) from point cloud to be registeredPSelecting n sampling points, wherein the distance between every two sampling points is greater than a preset minimum distance threshold value d in order to ensure that the sampled points have different FPFH characteristics as much as possible;
(2) in the target point cloudQMiddle search and point cloudPOne or more points with similar FPFH characteristics are sampled at the middle sampling point, and one point is randomly selected from the similar points to be used as a point cloudPIn the target point cloudQOne-to-one correspondence point of (1);
(3) and calculating a rigid body transformation matrix between the corresponding points, and then judging the performance of the current registration transformation result by solving the sum of the distance errors after the corresponding points are transformed.
Because the transformation matrix obtained by the sampling consistency initial registration algorithm is not accurate, the transformation matrix is only used for coarse registration, and the registration module in the PCL library can realize the sampling consistency initial registration algorithm.
And step three, performing accurate registration operation by using an ICP (inductively coupled plasma) algorithm.
The following conditions were set:
(1) setting source point cloudsPWith a target point cloudQIs 7, two points are considered as corresponding points (nearest neighbor points) as long as the two points are less than the threshold;
(2) setting a convergence judgment condition, a maximum allowable difference between two continuous conversions and a difference value between the last conversion and the current conversion;
(3) before the algorithm is considered to be converged, obtaining the difference value between the maximum allowable distance error and the error of two iterations;
(4) a maximum number of iterations 30 of the optimization run is set.
First, ICP registration of the M-plane and the R-plane is performed. Then, ICP registration of the M-plane and the L-plane is performed. And finally, performing ICP registration of the R surface and the L surface to realize 360-degree three-dimensional human head point cloud registration.
And fourthly, visually displaying the registration result of the 360-degree three-dimensional human head point cloud.
A point cloud visualization window is first created. Then setting is carried out, wherein the right side is represented by green point cloud, the middle side is represented by red point cloud, and the left side is represented by white point cloud. And splicing the point clouds which are registered pairwise together to realize 360-degree three-dimensional human head point cloud registration visualization.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (5)
1. A360-degree three-dimensional human head point cloud registration method based on PCL is characterized by comprising the following steps:
reading PCD files of three different visual angles of a human head, wherein the PCD files are L, M, R files and correspond to the left side, the middle side and the right side of the human head;
performing registration operation by using coarse registration based on sampling consistency to provide a better initial position for ICP fine registration;
performing accurate registration operation by using an ICP (inductively coupled plasma) algorithm to obtain a 360-degree three-dimensional human head point cloud registration result;
and fourthly, visually displaying the registration result of the 360-degree three-dimensional human head point cloud.
2. The PCL-based 360-degree three-dimensional human head point cloud registration method according to claim 1, characterized in that in the step one, PCD files of three different visual angles of the human head are read in, and the PCD files of three different visual angles of the human head are loaded respectively by using io: loadPCDFile < PointXYZ >.
3. The PCL-based 360-degree three-dimensional human head point cloud registration method according to claim 1, wherein in step two, a consistency initial registration algorithm is sampled, before executing the algorithm, a Fast Point Feature Histogram (FPFH) of the point cloud is calculated, and the algorithm flow is as follows:
step 2.1, from point cloud to be registeredPSelecting n sampling points, wherein the distance between every two sampling points is greater than a preset minimum distance threshold value d in order to ensure that the sampled points have different FPFH characteristics as much as possible;
step 2.2. in the target point cloudQMiddle search and point cloudPOne or more points with similar FPFH characteristics are sampled at the middle sampling point, and one point is randomly selected from the similar points to be used as a point cloudPIn the target point cloudQOne-to-one correspondence point of (1);
and 2.3, calculating a rigid body transformation matrix between the corresponding points, and then judging the performance of the current registration transformation result by solving the sum of the distance errors after the corresponding points are transformed.
4. The PCL-based 360-degree three-dimensional human head point cloud registration method according to claim 1, characterized in that in step three, an ICP algorithm is used to perform a precise registration operation, wherein the ICP registration operation comprises:
the ICP registration operation sets the following conditions: setting source point cloudsPWith a target point cloudQThe distance threshold of (2) is 7, and as long as the two points are smaller than the threshold, the two points are considered as corresponding points; setting a convergence judgment condition, a maximum allowable difference between two continuous conversions and a difference value between the last conversion and the current conversion; before the algorithm is considered to be converged, obtaining the difference value between the maximum allowable distance error and the error of two iterations; setting the maximum iteration number 30 of the optimized operation;
the ICP registration procedure is as follows: firstly, performing ICP registration of an M surface and an R surface; then, performing ICP registration of the M surface and the L surface; and finally, performing ICP registration of the R surface and the L surface to realize 360-degree three-dimensional human head point cloud registration.
5. The PCL-based 360-degree three-dimensional human head point cloud registration method according to claim 1, wherein the fourth step is specifically: firstly, creating a point cloud visualization window; then, setting is carried out, wherein the right side is represented by green point cloud, the middle side is represented by red point cloud, and the left side is represented by white point cloud; and finally, splicing the point clouds which are registered pairwise together to realize 360-degree three-dimensional human head point cloud registration visualization.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011614875.2A CN112561974A (en) | 2020-12-31 | 2020-12-31 | PCL-based 360-degree three-dimensional human head point cloud registration method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011614875.2A CN112561974A (en) | 2020-12-31 | 2020-12-31 | PCL-based 360-degree three-dimensional human head point cloud registration method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112561974A true CN112561974A (en) | 2021-03-26 |
Family
ID=75034771
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011614875.2A Pending CN112561974A (en) | 2020-12-31 | 2020-12-31 | PCL-based 360-degree three-dimensional human head point cloud registration method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112561974A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109816703A (en) * | 2017-11-21 | 2019-05-28 | 西安交通大学 | A kind of point cloud registration method based on camera calibration and ICP algorithm |
| CN110930456A (en) * | 2019-12-11 | 2020-03-27 | 北京工业大学 | Three-dimensional identification and positioning method of sheet metal part based on PCL point cloud library |
| CN111047631A (en) * | 2019-12-04 | 2020-04-21 | 广西大学 | Multi-view three-dimensional point cloud registration method based on single Kinect and round box |
-
2020
- 2020-12-31 CN CN202011614875.2A patent/CN112561974A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109816703A (en) * | 2017-11-21 | 2019-05-28 | 西安交通大学 | A kind of point cloud registration method based on camera calibration and ICP algorithm |
| CN111047631A (en) * | 2019-12-04 | 2020-04-21 | 广西大学 | Multi-view three-dimensional point cloud registration method based on single Kinect and round box |
| CN110930456A (en) * | 2019-12-11 | 2020-03-27 | 北京工业大学 | Three-dimensional identification and positioning method of sheet metal part based on PCL point cloud library |
Non-Patent Citations (1)
| Title |
|---|
| 程亚丽: "三维激光扫描点云数据配准算法研究", 《中国优秀硕士学位论文全文数据库 (基础科学辑)》, pages 008 - 107 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110340891B (en) | Mechanical arm positioning and grabbing system and method based on point cloud template matching technology | |
| CN106709947B (en) | Three-dimensional human body rapid modeling system based on RGBD camera | |
| CN109544456B (en) | Panoramic environment sensing method based on two-dimensional image and three-dimensional point cloud data fusion | |
| Pătrăucean et al. | State of research in automatic as-built modelling | |
| EP2026279B1 (en) | Method and system for aligning three-dimensional surfaces | |
| CN111340797A (en) | Laser radar and binocular camera data fusion detection method and system | |
| CN113178009B (en) | Indoor three-dimensional reconstruction method utilizing point cloud segmentation and grid repair | |
| CN107862735B (en) | RGBD three-dimensional scene reconstruction method based on structural information | |
| CN108830888B (en) | Coarse matching method based on improved multi-scale covariance matrix characteristic descriptor | |
| CN111179321A (en) | Point cloud registration method based on template matching | |
| CN112630469A (en) | Three-dimensional detection method based on structured light and multi-light-field camera | |
| He et al. | Linear approach for initial recovery of the exterior orientation parameters of randomly captured images by low-cost mobile mapping systems | |
| CN115546116A (en) | Method and system for extracting and calculating spacing of discontinuous surface of fully-covered rock mass | |
| Wang et al. | Point based registration of terrestrial laser data using intensity and geometry features | |
| CN111325828A (en) | Three-dimensional face acquisition method and device based on three-eye camera | |
| Li et al. | Feature point extraction and tracking based on a local adaptive threshold | |
| CN110969650B (en) | Intensity image and texture sequence registration method based on central projection | |
| CN112561974A (en) | PCL-based 360-degree three-dimensional human head point cloud registration method | |
| CN114897974B (en) | Target object space positioning method, system, storage medium and computer equipment | |
| Sumi et al. | Multiple TLS point cloud registration based on point projection images | |
| Wan et al. | A performance comparison of feature detectors for planetary rover mapping and localization | |
| CN112200850B (en) | ORB extraction method based on mature characteristic points | |
| Cai et al. | Automatic curve selection for lens distortion correction using Hough transform energy | |
| He et al. | Automatic docking recognition and location algorithm of port oil loading arm based on 3D laser point cloud | |
| CN109685837A (en) | A kind of heterologous remote sensing image matching method based on ground object structure similarity |
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 |