CN112344852A - Structured light three-dimensional scanning method - Google Patents

Abstract

The invention provides a structured light three-dimensional scanning method, which comprises the following steps: s1, calibrating the camera device, and establishing the corresponding relation between the image pixel coordinate and the world coordinate system; s2, generating compact point cloud data based on a Gray code combined with a linear motion coding structure optical positioning system; s3, inputting the point cloud data of the step S2 into a filter, and acquiring first point cloud data after noise reduction; s4, after rough alignment is carried out on the first point cloud data in the step S3, fine alignment is carried out to obtain second point cloud data; s5 the second point cloud data is processed by surface smoothing, then plane reconstruction is carried out, therefore, the structured light three-dimensional scanning method has the advantages of simple coding and decoding, strong anti-interference capability and high resolution, and is suitable for three-dimensional scanning of static targets.

Description

Structured light three-dimensional scanning method

Technical Field

The invention relates to a three-dimensional scanning scheme, in particular to a method for realizing three-dimensional scanning by adopting a structured light technology.

Background

Structured light is an important optical method of 3D scanning, which projects a set of mathematically structured light patterns that sequentially illuminate the object being measured. A camera, whose distance to the projector is known, captures images of a group of illuminated objects simultaneously. The pattern seen by the camera is distorted by the topography of the scanned object relative to a planar reference surface for calibration. The principle of geometric triangulation makes it possible to calculate the XYZ coordinates of each point on the surface of the scanned object. The obtained point cloud data is then used for the computational construction of a detailed 3D model of the scanned object surface.

Currently mainstream coding schemes are mainly divided into three major categories: time multiplex coding scheme, space domain coding side direct coding scheme

A time-multiplexed encoding scheme projects a series of patterns onto a scanned target surface at different times. For any point on the target surface, the different stripes projected to that point at different times constitute the encoded value for that point. The time-multiplexed coding scheme mainly includes the following sub-classes: binary-based coding, n-ary-based coding, gray code and line shift combined coding, and the like.

A spatial neighborhood coding scheme. The scheme only projects a structural light stripe pattern, and the coding value of each point on the target surface is determined by stripe information (color, brightness and the like) projected to the point and the neighborhood of the point, so that errors are easy to generate in the decoding process. Spatial neighborhood coding schemes are generally only suitable for scanning moving objects or objects with a relatively flat surface.

Direct coding schemes code each point of the structured light pattern and therefore require a very large number of colors, either reducing the resolution of the scan or periodically recycling the codewords. The scheme theoretically has extremely high scanning density because each point is coded, but the scheme is more susceptible to noise interference and influences on scanning precision in the decoding process because the distance between coded code words is reduced along with the increase of the number of code words.

The inventor researches a plurality of three-dimensional scanning coding schemes, and develops a structured light coding scheme based on a Gray code combined line shift scheme by comparing and comparing the characteristics and the limits of different coding schemes. The scheme has the advantages of simple and convenient coding and decoding, strong anti-interference capability and high resolution, and is suitable for three-dimensional scanning of static targets.

Disclosure of Invention

The invention mainly aims to provide a structured light three-dimensional scanning method, which is based on a structured light coding technology combining Gray codes with a linear shift scheme, so that three-dimensional scanning coding is simpler and more convenient, has strong anti-interference capability and high resolution, and is suitable for three-dimensional scanning of a static target.

In order to achieve the above object, the present invention provides a structured light three-dimensional scanning method, comprising the steps of: s1, calibrating the camera device, and establishing the corresponding relation between the image pixel coordinate and the world coordinate system; s2, generating compact point cloud data based on a Gray code combined with a linear motion coding structure optical positioning system; s3, inputting the point cloud data of the step S2 into a filter, and acquiring first point cloud data after noise reduction; s4, after rough alignment is carried out on the first point cloud data in the step S3, fine alignment is carried out to obtain second point cloud data; and S5, performing surface smoothing treatment on the second point cloud data, and performing plane reconstruction.

In a possible preferred alternative embodiment, the S1 step includes: and adopting a black and white checkerboard grid pattern with known three-dimensional size as a calibration reference.

In a possible preferred alternative embodiment, the S3 filter denoising step includes: aiming at the point cloud data of the background, filtering by adopting a radial filter; and for point cloud data derived from random noise, a statistical filter is adopted to filter out the random noise.

In a possible preferred alternative embodiment, the S4 rough alignment step includes: performing data preprocessing on the first point cloud data to find feature points, and calculating a geometric feature descriptor and calculating a corresponding relation of the geometric feature descriptor; after the corresponding relation of the descriptors with errors is eliminated: and calculating a coordinate transfer matrix.

In a possible preferred alternative embodiment, the S4 fine alignment step includes: a1, selecting an alignment point from the first point cloud for the first point cloud data after rough alignment; a2 finding the corresponding point of the selected point in A1 in the second point cloud according to the principle of the nearest adjacent point; a3 calculates the coordinate transfer matrix according to the points found in A1 and A2 and the corresponding relation of the points; a4 applying a coordinate transfer matrix to the first point cloud; a5, checking the coincidence degree of the first and second point clouds, if not reaching the preset standard, repeating the previous steps until the first and second point clouds are coincident.

The structured light three-dimensional scanning method provided by the invention has the advantages of simple and convenient coding and decoding, strong anti-interference capability and high resolution, and is suitable for three-dimensional scanning of a static target.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the invention and, together with the description, serve to explain the invention and not to limit the invention. In the drawings:

FIG. 1 is a schematic flow chart of a structured light three-dimensional scanning method according to the present invention;

fig. 2 is a schematic step diagram of a structured light three-dimensional scanning method according to the present invention.

Detailed Description

It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present invention will be described in detail below with reference to the embodiments with reference to the attached drawings.

In order to make those skilled in the art better understand the technical solution of the present invention, the technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, shall fall within the scope of the present invention.

It should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," as well as any variations thereof, are intended to cover non-exclusive inclusions.

Referring to fig. 1 to 2, according to an embodiment of the present invention, a structured light three-dimensional scanning method is provided, including: s1, calibrating the camera device, and establishing the corresponding relation between the image pixel coordinate and the world coordinate system. Specifically, the step S1 of the structured light three-dimensional scanning method is mainly used for calibrating the camera, and specifically, before performing three-dimensional scanning, the camera needs to be calibrated to establish the correspondence between the image pixel coordinate system and the world coordinate system. Generally, during the calibration process of a camera, a plurality of images of a calibration reference object are required to be shot, and the calibration reference object requires that three-dimensional features are known and has feature points which are easy to identify. Each picture taken provides a set of two-dimensional image coordinate system-three-dimensional world coordinate system correspondences. Therefore, only the parameters of each camera model are needed to be optimized to enable the two-dimensional-three-dimensional corresponding relation predicted by the model to be consistent with the corresponding relation in the actual picture as much as possible.

Meanwhile, the invention preferably selects a calibration method based on checkerboard calibration. The calibration method adopts the known three-dimensional black and white checkerboard pattern as a calibration reference, and has the advantages of convenient configuration, simple calibration process and accurate calculation.

S2, based on Gray code combined with line shift coding structure light positioning system, generating dense point cloud data. Specifically, in step S2, the present invention preferably uses a scheme based on gray code combined with line shift to generate dense point clouds, but due to the limitation of viewing angle, there may be a large gap between the generated point clouds that is not successfully sampled. In general, to enable a structured light based three-dimensional scanning system to successfully acquire three-dimensional information of a point on a target surface, the point needs to be covered by a structured light pattern projected via a projector, and the point needs to be captured by a camera. Thus, in general, only a portion of the point cloud of the surface of the target oriented in the direction of the projector and camera can be acquired in a single scan. To obtain full-angle three-dimensional structural information of a scanned object, the object must be scanned from a number of different angles.

S3, inputting the point cloud data of the step S2 into a filter, and acquiring the first point cloud data after noise reduction. Specifically, the point cloud obtained from the three-dimensional scan generally has some other point cloud data besides the three-dimensional point cloud data of the target, including but not limited to background point cloud data, random noise. The existence of the point cloud data causes great interference to subsequent point cloud operation, and therefore needs to be removed in advance.

For the point cloud data of the background, a certain separation distance exists between the target and the background in the scanning process, so that the inventor designs that the filtering is carried out by adopting a radial filter. For point cloud data derived from random noise, since the point cloud density is generally much lower than that of the target point cloud data, the average distance between adjacent points in these point clouds is also larger than that of the target point cloud. Based on this characteristic, the inventors have designed to use statistical filters to filter out these random noises.

And S4, after rough alignment is carried out on the first point cloud data in the step S3, fine alignment is carried out to obtain second point cloud data. Particularly, the problem of how to stitch a plurality of point clouds scanned from different angles into a complete target three-dimensional structure is a core problem in three-dimensional reconstruction, which is called the stitching of the point clouds. Because the performance of the algorithm in the prior art does not reach a mature step at present, splicing cannot be completed in one step, and based on the condition, the inventor designs a two-step alignment method.

Rough alignment:

(1) data pre-processing

(2) Finding feature points

(3) Computing geometric feature descriptors

(4) Calculating geometric feature descriptor correspondences

(5) Descriptor mapping to exclude errors

(6) Computing coordinate transfer matrices

And obtaining a rough coordinate transfer matrix between the two point clouds through the rough alignment step.

Fine alignment:

(1) points are selected from the first point cloud.

(2) And searching corresponding points of the points selected in the first step in the second point cloud according to the principle of the nearest adjacent points.

(3) And calculating a coordinate transfer matrix according to the points found in the first two steps and the corresponding relation of the points.

(4) A coordinate transfer matrix is applied to the first point cloud.

(5) And (4) checking the coincidence degree of the two point clouds, and repeating the previous steps if the coincidence degree of the two point clouds does not reach the preset standard until the point clouds are in accordance with the preset standard.

And S5, performing surface smoothing treatment on the second point cloud data, and performing plane reconstruction, thereby obtaining accurate three-dimensional point cloud model data.

In summary, the structured light three-dimensional scanning method provided by the invention has the advantages of simple and convenient encoding and decoding, strong anti-interference capability and high resolution, and is suitable for three-dimensional scanning of static targets.

The preferred embodiments of the invention disclosed above are intended to be illustrative only. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise embodiments disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention. The invention is limited only by the claims and the full scope and equivalents thereof, and any modification, equivalent replacement, or improvement made within the spirit and principle of the invention should be included in the protection scope of the invention.

Those skilled in the art will understand that all or part of the steps in the method according to the above embodiments may be implemented by a program, which is stored in a storage medium and includes several instructions to enable a single chip, a chip, or a processor (processor) to execute all or part of the steps in the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

In addition, any combination of various different implementation manners of the embodiments of the present invention is also possible, and the embodiments of the present invention should be considered as disclosed in the embodiments of the present invention as long as the combination does not depart from the spirit of the embodiments of the present invention.

Claims (5)

1. A structured light three-dimensional scanning method comprises the following steps:

s1, calibrating the camera device, and establishing the corresponding relation between the image pixel coordinate and the world coordinate system;

s2, generating compact point cloud data based on a Gray code combined with a linear motion coding structure optical positioning system;

s3, inputting the point cloud data of the step S2 into a filter, and acquiring first point cloud data after noise reduction;

s4, after rough alignment is carried out on the first point cloud data in the step S3, fine alignment is carried out to obtain second point cloud data;

and S5, performing surface smoothing treatment on the second point cloud data, and performing plane reconstruction.

2. The structured light three-dimensional scanning method according to claim 1, wherein the step of S1 comprises: and adopting a black and white checkerboard grid pattern with known three-dimensional size as a calibration reference.

3. The structured light three-dimensional scanning method according to claim 1, wherein the S3 filter denoising step comprises: aiming at the point cloud data of the background, filtering by adopting a radial filter; and for point cloud data derived from random noise, a statistical filter is adopted to filter out the random noise.

4. The structured light three-dimensional scanning method according to claim 1, wherein the S4 rough alignment step comprises: performing data preprocessing on the first point cloud data to find feature points, and calculating a geometric feature descriptor and calculating a corresponding relation of the geometric feature descriptor; after the corresponding relation of the descriptors with errors is eliminated: and calculating a coordinate transfer matrix.

5. The structured light three-dimensional scanning method according to claim 1, wherein the S4 fine alignment step comprises:

a1, selecting an alignment point from the first point cloud for the first point cloud data after rough alignment;

a2 finding the corresponding point of the point selected in the step A1 in the second point cloud according to the principle of the nearest point;

a3 calculating a coordinate transfer matrix according to the points found in the steps A1 and A2 and the corresponding relations of the points;

a4 applying a coordinate transfer matrix to the first point cloud;

a5, checking the coincidence degree of the first and second point clouds, if not reaching the preset standard, repeating the previous steps until the first and second point clouds are coincident.

Images