WO2021017471A1

WO2021017471A1 - Point cloud filtering method based on image processing, apparatus, and storage medium

Info

Publication number: WO2021017471A1
Application number: PCT/CN2020/078288
Authority: WO
Inventors: 张建民; 龙佳乐; 陈富健
Original assignee: 五邑大学
Priority date: 2019-07-30
Filing date: 2020-03-06
Publication date: 2021-02-04
Also published as: CN110458772B; CN110458772A

Abstract

A point cloud filtering method based on image processing, an apparatus, and a storage medium. The scattered points of an obtained initial point cloud are filtered to avoid the impact of the scattered points on subsequent processing; after initial point cloud coordinates corresponding to the filtered initial point cloud are uniquely numbered, because the point cloud coordinates are mapped to an image coordinate system, and the values of the initial point cloud coordinates are usually a floating point type, the data type of the initial point cloud coordinates is converted to an integer type to make mapping more accurate; in addition, after the connected components are extracted from a mapped image, the connected domain with the most integer-type point cloud coordinates corresponding to image coordinates in connected domains is retained, that is, the unique numbers of the initial point cloud coordinates corresponding to the image coordinates in this connected domain are retained; at this time, the filtered-out corresponding data in the other connected domains is noise data. The point cloud restored with the unique numbers is a target point cloud after filtering. There is no need to compare coordinates one by one, the amount of calculation is small, and point cloud noise is quickly filtered out.

Description

Point cloud filtering method, device and storage medium based on image processing

Technical field

The present invention relates to the technical field of data processing, in particular to a point cloud filtering method, device and storage medium based on image processing.

Background technique

In the fields of reconstruction of ancient buildings, reverse engineering of workpieces or survey engineering, non-contact three-dimensional measurement equipment can obtain the topography information of objects without touching the object, and it is widely used. Although the scanning accuracy has been steadily improved, It is inevitable that there will be problems such as lens distortion and inaccurate camera parameter estimation, resulting in the acquired topography information containing a lot of noise. The topography information is usually presented in the form of point clouds, so it is necessary to denoise the point clouds to ensure subsequent work Accuracy. Existing solutions usually process point clouds based on three-dimensional space, and remove noise by calculating the relationship between the initial point clouds. However, the initial point cloud acquired by the three-dimensional measuring equipment often has a huge amount of data, and the direct processing of the initial point cloud has a large amount of calculation. , The shortcomings of low efficiency and high complexity.

Summary of the invention

In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide a point cloud filtering method, device and storage medium based on image processing, reducing the amount of point cloud denoising calculations, and completing point cloud filtering quickly and efficiently.

The technical solution adopted by the present invention to solve its problems is: In the first aspect, the present invention provides a point cloud filtering method based on image processing, which includes the following steps:

The client obtains the initial point cloud, and filters the scattered points of the initial point cloud;

The client obtains the initial point cloud coordinates corresponding to the initial point cloud in the point cloud coordinate system, and numbers the initial point cloud coordinates to obtain a coordinate number uniquely corresponding to the initial point cloud coordinates;

The client converts the initial point cloud coordinates into integer point cloud coordinates whose data type is integer, establishes an image coordinate system, and maps the integer point cloud coordinates to the image coordinate system;

The client acquires a mapped image composed of data in the image coordinate system, and after extracting connected components from the mapped image, sets the connected domain with the most integer point cloud coordinates corresponding to the image coordinates as the target connected Domain, and filter out other connected domains;

Acquire the coordinate numbers included in the target connected domain, restore the integer point cloud coordinates to the corresponding initial point cloud coordinates according to the coordinate numbers, and map them to the point cloud coordinate system to complete point cloud filtering .

Further, the scattered points in the initial point cloud are filtered by the k nearest neighbor algorithm.

Further, the image coordinate system is established according to the origin, x-axis and y-axis of the point cloud coordinate system.

Further, the coordinate value in the integer point cloud coordinate is a value obtained by rounding the coordinate value of the initial point cloud coordinate.

Further, before extracting the connected components from the mapped image, the method further includes: performing a closing operation on the mapped image to bridge the broken area.

In a second aspect, the present invention provides a device for performing a point cloud filtering method based on image processing, including a CPU unit configured to perform the following steps:

Further, the CPU unit is further configured to perform the following steps: performing a closing operation on the mapped image to bridge the fractured area.

In a third aspect, the present invention provides a device for performing a point cloud filtering method based on image processing, including at least one control processor and a memory for communicating with the at least one control processor; the memory stores at least An instruction executed by a control processor, the instruction is executed by at least one control processor, so that the at least one control processor can execute the above-mentioned point cloud filtering method based on image processing.

In a fourth aspect, the present invention provides a computer-readable storage medium, the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to make a computer execute the above-mentioned image processing-based point cloud filtering method.

In a fifth aspect, the present invention also provides a computer program product, the computer program product includes a computer program stored on a computer-readable storage medium, the computer program includes program instructions, when the program instructions are executed by a computer , Let the computer execute the point cloud filtering method based on image processing as described above.

The one or more technical solutions provided in the embodiments of the present invention have at least the following beneficial effects: the present invention provides a point cloud filtering method, device and storage medium based on image processing, which scatters the obtained initial point cloud Filter to avoid the impact of scattered points on subsequent processing; after the initial point cloud coordinates corresponding to the filtered initial point cloud are uniquely numbered, the point cloud coordinates are mapped to the image coordinate system, and the initial point cloud coordinates are usually It is a floating point type, so the data type of the initial point cloud coordinates is converted to an integer type to make the mapping more accurate; at the same time, after the connected components of the mapped image are extracted, the image coordinates in the connected domain are the most corresponding integer point cloud coordinates. The connected domain of, which retains the unique number of the initial point cloud coordinates corresponding to the image coordinates in this connected domain. At this time, the corresponding data in other connected domains filtered out is noise data, and the point cloud restored by the unique number is the filtered The target point cloud does not need to be compared one by one, the calculation amount is small, and the point cloud noise is quickly filtered out.

Description of the drawings

The present invention will be further explained below with reference to the drawings and examples.

Fig. 1 is a flowchart of a method according to an embodiment of the present invention;

Fig. 2 is an example diagram of an initial point cloud according to an embodiment of the present invention;

3 is an example diagram of the effect of filtering scattered points in the initial point cloud according to an embodiment of the present invention;

FIG. 4 is an example diagram of the effect of a mapped image according to an embodiment of the present invention;

FIG. 5 is an example diagram of the effect of a closed operation on a mapped image according to an embodiment of the present invention;

FIG. 6 is an example diagram of the effect of acquiring target connected domains according to an embodiment of the present invention;

FIG. 7 is a diagram showing an example of the effect of point cloud filtering in an embodiment of the present invention;

Fig. 8 is a schematic diagram of an apparatus for executing a point cloud filtering method based on image processing according to an embodiment of the present invention.

Detailed ways

In order to make the objectives, technical solutions, and advantages of the present invention clearer, the following further describes the present invention in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, but not to limit the present invention.

It should be noted that if there is no conflict, the various features in the embodiments of the present invention can be combined with each other, and all fall within the protection scope of the present invention. In addition, although functional modules are divided in the schematic diagram of the device, and the logical sequence is shown in the flowchart, in some cases, it may be different from the module division in the device, or the sequence shown in the flowchart may be executed. Or the steps described.

It should be noted that the data of the present invention can be collected by common collection equipment on the market. The present invention does not involve the improvement of the collection equipment, but only processes the data acquired by the collection equipment.

Referring to Fig. 1 to Fig. 7, the first embodiment of the present invention provides a point cloud filtering method based on image processing, including the following steps:

Step S100, the client obtains the initial point cloud, and filters scattered points of the initial point cloud;

Step S200: The client obtains the initial point cloud coordinates corresponding to the initial point cloud in the point cloud coordinate system, and numbers the initial point cloud coordinates to obtain a coordinate number uniquely corresponding to the initial point cloud coordinates;

Step S300: The client converts the initial point cloud coordinates into integer point cloud coordinates whose data type is integer, establishes an image coordinate system, and maps the integer point cloud coordinates to the image coordinate system ；

Step S400: The client acquires a mapped image composed of data in the image coordinate system, and after extracting connected components from the mapped image, sets the image coordinates corresponding to the connected domain with the most integer point cloud coordinates Is the target connected domain, and filter out other connected domains;

Step S500: Obtain the coordinate numbers included in the target connected domain, restore the integer point cloud coordinates to the corresponding initial point cloud coordinates according to the coordinate numbers, and map them to the point cloud coordinate system to complete Point cloud filtering.

Among them, it should be noted that there are many scattered points in the point cloud obtained by a three-dimensional device. For example, the dots shown in Figure 2 cause many scattered points. If it is not filtered out, it is easy to cause the target point cloud. Large interference. Therefore, in this embodiment, it is preferable to filter out scattered points first after acquiring the initial point cloud, which can eliminate the interference of scattered points. An example diagram obtained after filtering is shown in FIG. 3.

Among them, it should be noted that the initial point cloud coordinates are three-dimensional coordinates, that is, including the x-axis, y-axis, and z-axis. Therefore, the initial point cloud coordinates in this embodiment can be expressed as (x, y, z), then step S200 After the initial point cloud coordinates are numbered, the initial point cloud coordinates are (x _i , y _i , z _i ), and its unique coordinate number is i. The specific value of i is determined by the order of the initial point cloud coordinates, for example, according to x When the axis values are arranged in order, the coordinate with the smallest x-axis value is x ₁ , and so on, so I won’t repeat them here.

Among them, in this embodiment, due to the accuracy requirements of the acquisition device, the data type of the initial point cloud usually obtained is floating point. If mapping is to be performed, the floating point data needs to be converted into integer data. This embodiment The rounding method is preferred to obtain integer data more easily.

Among them, it should be noted that after the mapping is completed in step S300 and the floating-point point cloud coordinates are converted into integer coordinates, there is a situation where one or more floating-point coordinates correspond to an integer coordinate, that is, an integer point When cloud coordinates are mapped to image coordinates, there will be one or more integer point cloud coordinates corresponding to one image coordinate, and the effect diagram is shown in Figure 4. However, the coordinate number corresponding to the integer point cloud coordinate is unique, and the image coordinate also corresponds to the integer point cloud coordinate and its corresponding coordinate number. Therefore, it is possible to distinguish the integer point cloud coordinates corresponding to the image coordinate.

Among them, it should be noted that in this embodiment, preferably, in step S400, the number of integer point cloud coordinates corresponding to the image coordinates in each connected domain is counted, and the connected domain with the largest number of integer point cloud coordinates corresponding to the image coordinates in the connected domain is reserved. , That is, the coordinate number corresponding to the integral point cloud coordinate of the image coordinate in this connected domain is retained, and other connected domains are filtered out, that is, the coordinate number corresponding to the integral point cloud coordinate of the image coordinate in other connected domains is filtered out, and The effect example is shown in Figure 6.

Among them, it should be noted that the number corresponding to the integer point cloud coordinates retained in step S400 is restored to the corresponding initial point cloud coordinates, that is, the integer point cloud coordinates with inconsistent coordinate numbers in the connected domain are filtered out, thereby achieving Filter out the noise point cloud, and only keep the target point cloud. Because each initial point cloud coordinate is numbered, the corresponding initial point cloud coordinate is found in the initial point cloud according to the reserved number in step S400, and point cloud filtering is realized. The effect example is shown in Figure 7. .

Further, in another embodiment of the present invention, the scattered points in the initial point cloud are filtered by the k nearest neighbor algorithm.

Among them, it should be noted that the filtering of scattered points can be completed by any algorithm to achieve the effect of filtering scattered points. This embodiment preferably adopts the k nearest neighbor algorithm to achieve efficient and accurate filtering of scattered points. The algorithm can be selected according to actual needs.

Further, in another embodiment of the present invention, the image coordinate system is established according to the origin, x-axis and y-axis of the point cloud coordinate system.

It should be noted that, in this embodiment, it is also preferable that the length H and the width W of the image coordinate system respectively satisfy the following formulas:

H=max(y _k )-min(y _k ); k=1, 2, 3..., n;

W=max(x _k )-min(x _k ); k=1, 2, 3..., n.

Among them, x _k is the abscissa value of the integer point cloud coordinate converted from the initial point cloud coordinate with the coordinate number k minus the minimum coordinate value, and y _k is the initial point cloud coordinate with the coordinate number k minus the minimum coordinate value The ordinate value of the integer point cloud coordinate after conversion.

Further, in another embodiment of the present invention, the coordinate value in the integer point cloud coordinate is a value obtained by rounding the coordinate value of the initial point cloud coordinate.

Among them, it should be noted that rounding is the preferred embodiment of this embodiment, and methods such as rounding up or rounding down can also be used to obtain integer data, and the specific method can be adjusted according to actual needs.

Referring to FIG. 5, further, in another embodiment of the present invention, before extracting connected components from the mapped image, the method further includes: performing a closing operation on the mapped image to bridge the fractured area.

Among them, since there are many fracture areas in the mapped image, in this embodiment, the closed operation is preferably used to bridge the fractured areas to improve the accuracy of the image. The schematic diagram of the effect is shown in Figure 5, and it should be noted that the closed The operation is an algorithm in the prior art, and this embodiment does not involve the improvement of the closed operation algorithm, and will not be repeated here.

Referring to Figure 8, the second embodiment of the present invention also provides a device for performing a point cloud filtering method based on image processing. The device is a smart device, such as a smart phone, a computer, and a tablet. Take the computer as an example.

The computer 8000 for performing a point cloud filtering method based on image processing includes a CPU unit 8100, and the CPU unit 8100 is configured to perform the following steps:

The client acquires the initial point cloud coordinates corresponding to the initial point cloud in the point cloud coordinate system, and numbers the initial point cloud coordinates to obtain a coordinate number uniquely corresponding to the initial point cloud coordinates;

Further, in another embodiment of the present invention, the CPU unit 8100 is further configured to perform the following steps: performing a closing operation on the mapped image to bridge the fractured area.

The computer 8000 and the CPU unit 8100 can be connected by a bus or other means. The computer 8000 also includes a memory. As a non-transitory computer-readable storage medium, the memory can be used to store non-transitory software programs and non-transitory Computer-executable programs and modules, such as program instructions/modules corresponding to the device for executing the point cloud filtering method based on image processing in the embodiment of the present invention. The computer 8000 runs the non-transitory software programs, instructions, and modules stored in the memory to control the CPU unit 8100 to execute various functional applications and data processing for performing the point cloud filtering method based on image processing, that is, to implement the above method Example point cloud filtering method based on image processing.

The memory may include a storage program area and a storage data area. The storage program area may store an operating system and an application program required by at least one function; the storage data area may store data created according to the use of the CPU unit 8100 and the like. In addition, the memory may include a high-speed random access memory, and may also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid state storage devices. In some embodiments, the memory may optionally include a memory remotely provided with respect to the CPU unit 8100, and these remote memories may be connected to the computer 8000 via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

The one or more modules are stored in the memory, and when executed by the CPU unit 8100, the point cloud filtering method based on image processing in the foregoing method embodiment is executed.

The embodiment of the present invention also provides a computer-readable storage medium that stores computer-executable instructions that are executed by the CPU unit 8100 to implement the aforementioned image processing-based points. Cloud filtering method.

The device embodiments described above are merely illustrative, and the devices described as separate components may or may not be physically separated, that is, they may be located in one place, or they may be distributed on multiple network devices. Some or all of the modules may be selected according to actual needs to achieve the objectives of the solutions of the embodiments.

It should be noted that since the device for performing the point cloud filtering method based on image processing in this embodiment is based on the same inventive concept as the above-mentioned point cloud filtering method based on image processing, the corresponding content in the method embodiment The same applies to the embodiments of the device, and will not be described in detail here.

Through the description of the above implementation manners, those skilled in the art can clearly understand that each implementation manner can be implemented by means of software plus a general hardware platform. Those skilled in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented by computer programs instructing relevant hardware. The programs can be stored in a computer readable storage medium. , May include the flow of the embodiment of the above method. Wherein, the storage medium may be a magnetic disk, an optical disc, a read-only memory (Read Only Memory, ROM), or a random access memory (Random Access Memory, RAM), etc.

The above is a detailed description of the preferred implementation of the present invention, but the present invention is not limited to the above-mentioned embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. Equivalent modifications or replacements are all included in the scope defined by the claims of this application.

Claims

A point cloud filtering method based on image processing, characterized by comprising the following steps: a client obtains an initial point cloud, and filters scattered points of the initial point cloud;

The client obtains the initial point cloud coordinates corresponding to the initial point cloud in the point cloud coordinate system, and numbers the initial point cloud coordinates to obtain a coordinate number uniquely corresponding to the initial point cloud coordinates;

The client converts the initial point cloud coordinates into integer point cloud coordinates whose data type is integer, establishes an image coordinate system, and maps the integer point cloud coordinates to the image coordinate system;

The client acquires a mapped image composed of data in the image coordinate system, and after extracting connected components from the mapped image, sets the connected domain with the most integer point cloud coordinates corresponding to the image coordinates as the target connected Domain, and filter out other connected domains;

Acquire the coordinate numbers included in the target connected domain, restore the integer point cloud coordinates to the corresponding initial point cloud coordinates according to the coordinate numbers, and map them to the point cloud coordinate system to complete point cloud filtering .
The point cloud filtering method based on image processing according to claim 1, wherein the scattered points in the initial point cloud are filtered by the k nearest neighbor algorithm.
A point cloud filtering method based on image processing according to claim 1, wherein the image coordinate system is established based on the origin, x-axis and y-axis of the point cloud coordinate system.
A point cloud filtering method based on image processing according to claim 1, characterized in that: the coordinate value in the integer point cloud coordinates is obtained by rounding the coordinate value of the initial point cloud coordinate value.
A point cloud filtering method based on image processing according to claim 1, characterized in that, before extracting connected components from the mapped image, it further comprises: performing a closing operation on the mapped image to bridge the fractured area .
A device for performing a point cloud filtering method based on image processing, characterized in that it comprises a CPU unit, and the CPU unit is configured to perform the following steps:

The client obtains the initial point cloud, and filters the scattered points of the initial point cloud;

The client obtains the initial point cloud coordinates corresponding to the initial point cloud in the point cloud coordinate system, and numbers the initial point cloud coordinates to obtain a coordinate number uniquely corresponding to the initial point cloud coordinates;

The client converts the initial point cloud coordinates into integer point cloud coordinates whose data type is integer, establishes an image coordinate system, and maps the integer point cloud coordinates to the image coordinate system;

The client acquires a mapped image composed of data in the image coordinate system, and after extracting connected components from the mapped image, sets the connected domain with the most integer point cloud coordinates corresponding to the image coordinates as the target connected Domain, and filter out other connected domains;

Acquire the coordinate numbers included in the target connected domain, restore the integer point cloud coordinates to the corresponding initial point cloud coordinates according to the coordinate numbers, and map them to the point cloud coordinate system to complete point cloud filtering .
The apparatus for performing a point cloud filtering method based on image processing according to claim 6, wherein the CPU unit is further configured to perform the following steps: closing the mapped image to bridge the broken area.
A computer-readable storage medium, characterized in that: the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to make a computer execute any one of claims 1-5 A point cloud filtering method based on image processing.