A method and device for extracting linear features from a three-dimensional point cloud

Abstract

The application relates to a three-dimensional point cloud linear feature extraction method and device, which comprises the following steps: obtaining a three-dimensional point cloud of a linear feature to be extracted; screening at least one target element point cloud from the three-dimensional point cloud based on a threshold segmentation algorithm; performing edge feature extraction on each target element point cloud and saving each target element point cloud after edge feature extraction; projecting each target element point cloud after edge feature extraction to a two-dimensional plane to obtain one or more plane views; and extracting at least one linear feature from the one or more plane views through an edge line detection algorithm. The application can effectively extract key features, reduce the amount of calculation and improve the accuracy by screening the point cloud through threshold segmentation and edge feature extraction.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of computer vision, and particularly relates to a three-dimensional point cloud line feature extraction method and device. BACKGROUND

[0002] With the continuous maturity of three-dimensional scanning technology, three-dimensional point cloud processing technology has gradually become an important research focus in the field of computer vision, and is widely applied in medical treatment, archaeology, artificial intelligence and various disciplines. Similarly, the production of automatic driving high-precision maps cannot be separated from three-dimensional scanning technology.

[0003] The original point cloud data obtained by three-dimensional scanning generally has the characteristics of complex data quantity, and the composition of the point cloud data does not have a relatively regular topological structure. If the line features can be extracted from the point cloud data to describe the shape, size and surface texture of the object, many features, it will enable people to more intuitively understand the geometric features of the object itself. For example, the ridge line extraction of valley topography can be used to analyze some topography in China. The extraction of facial feature can be used for face recognition. In high-precision map production, the line feature association between point cloud and image can be used for 3D-2D data registration.

[0004] In recent years, for the extraction of line features, most research results are concentrated on 2D images, including line extraction of a single picture and 3D line reconstruction of multi-view pictures. Only a few research results are directed to the extraction of line features from point clouds, and these methods often have problems such as complex extraction process, slow extraction speed, and poor extraction quality. SUMMARY

[0005] To solve the problems of complexity, slow speed and low precision of the existing three-dimensional point cloud line extraction method, in the first aspect of the present application, a three-dimensional point cloud line feature extraction method is provided, comprising: obtaining a three-dimensional point cloud of a line feature to be extracted; based on a threshold segmentation algorithm, screening at least one target element point cloud from the three-dimensional point cloud; performing edge feature extraction on each target element point cloud, and saving each target element point cloud after edge feature extraction; projecting each target element point cloud after edge feature extraction to a two-dimensional plane to obtain one or more plane views; and extracting at least one line feature from the one or more plane views by an edge line detection algorithm.

[0006] In some embodiments of the present application, the threshold segmentation algorithm is used to screen at least one target element point cloud from the three-dimensional point cloud, comprising: using an adaptive threshold segmentation algorithm to remove point clouds with laser reflection intensity lower than a threshold from the three-dimensional point cloud.

[0007] In some embodiments of the present application, the edge feature extraction on each target element point cloud comprises: performing edge feature extraction on each target element point cloud by a boundary estimation algorithm.

[0008] In some embodiments of the present application, the projecting each edge feature extracted target element point cloud to a two-dimensional plane to obtain one or more plane views comprises: projecting each edge feature extracted target element point cloud to a two-dimensional plane to obtain one or more plane views by a front view projection method.

[0009] In some embodiments of the present application, the extracting at least one straight line feature from the one or more plane views by an edge line detection algorithm comprises: extracting at least one straight line feature from the one or more plane views by a CannyLine algorithm.

[0010] In the above-mentioned embodiments, further comprising: extracting point cloud data corresponding to each straight line feature by a proximity search algorithm.

[0011] A second aspect of the present application provides a three-dimensional point cloud straight line feature extraction device, comprising: an acquisition module configured to acquire a three-dimensional point cloud to be extracted for a straight line feature; a threshold segmentation algorithm-based module configured to screen at least one target element point cloud from the three-dimensional point cloud; a first extraction module configured to perform edge feature extraction on each target element point cloud and save each edge feature extracted target element point cloud; a projection module configured to project each edge feature extracted target element point cloud to a two-dimensional plane to obtain one or more plane views; and a second extraction module configured to extract at least one straight line feature from the one or more plane views by an edge line detection algorithm.

[0012] Further comprising: a third extraction module configured to extract point cloud data corresponding to each straight line feature by a proximity search algorithm.

[0013] A third aspect of the present application provides an electronic device, comprising: one or more processors; and a storage device configured to store one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the three-dimensional point cloud straight line feature extraction method provided in the first aspect of the present application.

[0014] A fourth aspect of the present application provides a computer readable medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the three-dimensional point cloud straight line feature extraction method provided in the first aspect of the present application.

[0015] The present application has the following beneficial effects:

[0016] Compared with the prior art, ① the key features can be effectively extracted by screening the point cloud through threshold segmentation and edge feature, and the calculation workload is greatly reduced. ② The interference of the redundant feature lines in the road elements is removed, so that the line feature extraction accuracy is high. ③ Compared with the previous method, the process of the method is simple and easy to implement, and provides a good solution for the three-dimensional point cloud line feature extraction of high-precision map making. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 A basic flowchart of a three-dimensional point cloud line feature extraction method in some embodiments of the application is shown in the figure.

[0018] Figure 2 A schematic diagram of a road point cloud without edge feature extraction in some embodiments of the application is shown in the figure.

[0019] Figure 3 A schematic diagram of a road point cloud with edge feature extraction in some embodiments of the application is shown in the figure.

[0020] Figure 4 A schematic diagram of a point cloud front projection in some embodiments of the application is shown in the figure.

[0021] Figure 5 A schematic diagram of a point cloud front projection in some embodiments of the application is shown in the figure.

[0022] Figure 6 A structural schematic diagram of a three-dimensional point cloud line feature extraction device in some embodiments of the application is shown in the figure.

[0023] Figure 7 A structural schematic diagram of an electronic device in some embodiments of the application is shown in the figure. DETAILED DESCRIPTION

[0024] The principles and characteristics of the application are described below in conjunction with the drawings, and the examples are only used to explain the application and not to limit the scope of the application.

[0025] REFERENCE Figure 1 AND Figure 2 In the first aspect of the application, a three-dimensional point cloud line feature extraction method is provided, comprising: S100. obtaining a three-dimensional point cloud to be extracted for line features; based on a threshold segmentation algorithm, screening at least one target element point cloud from the three-dimensional point cloud; S200. edge feature extraction is performed on each target element point cloud, and each target element point cloud after edge feature extraction is saved; S300. Project each target element point cloud after edge feature extraction to a two-dimensional plane to obtain one or more plane views; S400. At least one line feature is extracted from the one or more plane views through an edge line detection algorithm.

[0026] REFERENCEFigure 2 With Figure 3 In step S100 of some embodiments of the present application, the filtering at least one target element point cloud from the three-dimensional point cloud based on the threshold segmentation algorithm comprises: removing the point cloud with laser reflection intensity lower than the threshold from the three-dimensional point cloud by an adaptive threshold segmentation algorithm. Most of the road surface information can be filtered out by threshold segmentation, and only the artificial marked element information is retained.

[0027] In step S200 of some embodiments of the present application, the edge feature extraction of each target element point cloud comprises: performing edge feature extraction on each target element point cloud by a boundary estimation algorithm.

[0028] It can be understood that the advantages of edge feature processing are: ① the point cloud inside the lane line, guardrail, arrow and other map elements can be removed, the subsequent calculation amount is reduced, and the running efficiency is improved. ② At the same time, the internal point cloud can be avoided to be extracted as a line feature by mistake, and the detection accuracy is reduced.

[0029] Reference Figure 4 With Figure 5 In step S300 of some embodiments of the present application, the projecting each target element point cloud after edge feature extraction to a two-dimensional plane to obtain one or more plane views comprises: projecting each target element point cloud after edge feature extraction to a two-dimensional plane to obtain one or more plane views by a front view projection method.

[0030] In step S400 of some embodiments of the present application, the extracting at least one straight line feature from the one or more plane views by an edge line detection algorithm comprises: extracting at least one straight line feature from the one or more plane views by a CannyLine algorithm.

[0031] In the above embodiments, it further comprises: extracting the point cloud data corresponding to each straight line feature by a proximity search algorithm. Optionally, the purpose of extracting the point cloud data corresponding to each straight line feature is achieved by a DBSCAN clustering method.

[0032] It can be understood that, after the point cloud preprocessing, threshold segmentation, point cloud edge feature extraction and other processing of the present application, the noise points and internal point clouds of objects can be removed, and only the edge feature information of the point cloud object is retained; the point cloud is saved in an image format by front view projection, the three-dimensional line feature problem is converted into a two-dimensional line feature problem; the accurate line feature information is extracted by using a CannyLine line feature detection algorithm; and finally the line feature point cloud data in the point cloud is extracted according to the line feature information by using a fast nearest neighbor search algorithm.

[0033] The application eliminates the interference of noise points and the interference of redundant point clouds in the object by combining the method of pretreatment, adaptive threshold segmentation and edge feature processing; the three-dimensional line feature extraction problem is converted into a two-dimensional line feature extraction problem by projection; compared with the previous method, the method has novel ideas, simple process and high extraction accuracy, and can provide a good solution for three-dimensional point cloud linear feature extraction for high-precision map making.

[0034] In one embodiment of the application, the following steps are included:

[0035] Step 1: using a vehicle-mounted laser radar device to obtain point cloud data of a road.

[0036] Step 2: pretreating the point cloud data to eliminate the point cloud below the road surface.

[0037] Step 3: using adaptive threshold segmentation (OTSU algorithm) to eliminate point clouds with low intensity, thereby extracting human-labeled road information (such as ground lane lines, arrows, guardrails, etc.).

[0038] Step 4: eliminating the interference of redundant point cloud data in the elements of the high-precision map by extracting the edge features of the point cloud.

[0039] Step 5: using forward projection to project the point cloud data to a 2D plane to generate a plane view.

[0040] Step 6: extracting the line features in the plane view by using the CannyLine line detection algorithm.

[0041] Step 7: extracting the point cloud data corresponding to the line features by using the fast nearest neighbor search algorithm.

[0042] Step 8: outputting and saving the line feature point cloud data.

[0043] It should be noted that although the embodiments of the application are based on the development environment of VS 2015, and by means of the two powerful open source libraries OPENCV and PCL, the point cloud data is pretreated, threshold segmented, edge feature extracted, point cloud forward projected, line feature detected and fast nearest neighbor searched to realize the extraction of the linear features of the three-dimensional point cloud data, but this does not affect the implementation of the above-mentioned point cloud feature extraction method by adapting or porting based on other development environments.

[0044] Embodiment 2

[0045] Reference Figure 6In a second aspect, the present application provides a device 1 for extracting linear features from a three-dimensional point cloud, comprising: an obtaining module 11 configured to obtain a three-dimensional point cloud from which linear features are to be extracted; a threshold segmentation algorithm configured to filter at least one target element point cloud from the three-dimensional point cloud; a first extraction module 12 configured to extract edge features from each target element point cloud and save each target element point cloud after edge feature extraction; a projection module 13 configured to project each target element point cloud after edge feature extraction onto a two-dimensional plane to obtain one or more plane views; and a second extraction module 14 configured to extract at least one linear feature from the one or more plane views by using an edge line detection algorithm.

[0046] Further, the device 1 further comprises: a third extraction module configured to extract point cloud data corresponding to each linear feature by using a proximity search algorithm.

[0047] Embodiment 3

[0048] Reference Figure 7 In a third aspect, the present application provides an electronic device, comprising: one or more processors; and a storage device storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method for extracting linear features from a three-dimensional point cloud according to the first aspect of the present application.

[0049] The electronic device 500 can include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 501 that can perform various appropriate actions and processes according to programs stored in a read-only memory (ROM) 502 or loaded into a random access memory (RAM) 503 from a storage device 508. Various programs and data required for the operation of the electronic device 500 are also stored in the RAM 503. The processing device 501, the ROM 502, and the RAM 503 are connected to each other through a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0050] Generally, the following devices can be connected to the I / O interface 505: input devices 506 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; output devices 507 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 508 including, for example, a hard disk, etc.; and communication devices 509. The communication devices 509 can allow the electronic device 500 to communicate wirelessly or via a wire with other devices to exchange data. Although Figure 7 The electronic device 500 is shown with various devices, but it should be understood that not all of the shown devices are required to be implemented or present. More or fewer devices can alternatively be implemented or present. Figure 7Each block in the flow diagrams of FIGS. 1-3 can represent a module, segment, or portion of code, which comprises one or more instructions that implement the specified logical functions. It should also be noted that each block of the flow diagrams and combinations thereof can represent a module, segment, or portion of code, which comprises one or more instructions that implement the specified logical functions.

[0051] In particular, according to embodiments of the present disclosure, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, embodiments of the present disclosure include a computer program product comprising a computer program carried on a computer readable medium, the computer program comprising program code for performing the methods illustrated by the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via the communication device 509, or installed from the storage device 508, or installed from the ROM 502. When the computer program is executed by the processing device 501, the above-mentioned functions defined in the methods of embodiments of the present disclosure are performed. It should be noted that the computer readable medium described in embodiments of the present disclosure can be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or apparatus, or any suitable combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In embodiments of the present disclosure, the computer readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device. In embodiments of the present disclosure, the computer readable signal medium can include a data signal carried in a baseband or as part of a carrier wave, in which the computer readable program code is carried. Such a propagated data signal can take any of a variety of forms, including but not limited to electro-magnetic, optical, or any suitable combination thereof. The computer readable signal medium can also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer readable medium can be transmitted by any suitable medium, including but not limited to wire, cable, RF, etc., or any suitable combination of the above.

[0052] The computer readable medium described above can be included in the electronic device described above; or can exist separately from the electronic device and be accessed via the electronic device. The computer readable medium carries one or more computer programs that, when executed by the electronic device, cause the electronic device to:

[0053] Computer program code for carrying out operations of embodiments of the present disclosure can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++, Python, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0054] The computer program instructions can also be loaded onto a computer or other programmable information processing apparatus to cause a series of operations to be performed on the computer or other programmable information processing apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable information processing apparatus implement the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0055] The above descriptions are only the preferred embodiment of the application, not intended to limit the application and any modification, equivalent replacement and improvement made within the spirit and principle of the application shall be included in the protection scope of the application.

Claims

1. A method for extracting straight line features from a three-dimensional point cloud, characterized in that, include: Obtain the 3D point cloud of the line features to be extracted; Based on the threshold segmentation algorithm, at least one target feature point cloud is selected from the three-dimensional point cloud: the laser reflection intensity is removed from the three-dimensional point cloud by the adaptive threshold segmentation algorithm. Edge features are extracted from the point cloud of each target element, and the point cloud of each target element after edge feature extraction is saved. Project the point cloud of each target feature extracted from the edge features onto a two-dimensional plane to obtain one or more planar views; At least one straight line feature is extracted from the one or more planar views using an edge line detection algorithm.

2. The method for extracting straight line features from three-dimensional point clouds according to claim 1, characterized in that, The step of extracting edge features for each target feature point cloud includes: extracting edge features for each target feature point cloud using a boundary estimation algorithm.

3. The method for extracting straight line features from three-dimensional point clouds according to claim 1, characterized in that, The step of projecting the point cloud of each target feature extracted from its edge features onto a two-dimensional plane to obtain one or more planar views includes: By using the forward projection method, the point cloud of each target feature after edge feature extraction is projected onto a two-dimensional plane to obtain one or more planar views.

4. The method for extracting straight line features from three-dimensional point clouds according to claim 1, characterized in that, Extracting at least one straight line feature from the one or more planar views using an edge line detection algorithm includes: Extract at least one straight line feature from the one or more planar views using the CannyLine algorithm.

5. The method for extracting straight line features from a three-dimensional point cloud according to any one of claims 1 to 4, characterized in that, Also includes: Point cloud data corresponding to each straight line feature is extracted using a nearest neighbor search algorithm.

6. A three-dimensional point cloud linear feature extraction device, characterized in that, include: The acquisition module is used to acquire the 3D point cloud of the straight line features to be extracted; Based on the threshold segmentation algorithm, at least one target feature point cloud is selected from the three-dimensional point cloud; The first extraction module is used to obtain the 3D point cloud of the straight line features to be extracted; Based on the threshold segmentation algorithm, at least one target feature point cloud is selected from the three-dimensional point cloud: the laser reflection intensity is removed from the three-dimensional point cloud by the adaptive threshold segmentation algorithm. The projection module is used to project the point cloud of each target feature extracted from the edge features onto a two-dimensional plane to obtain one or more planar views. The second extraction module is used to extract at least one straight line feature from the one or more planar views using an edge line detection algorithm.

7. The three-dimensional point cloud straight line feature extraction device according to claim 6, characterized in that, Also includes: The third extraction module is used to extract the point cloud data corresponding to each straight line feature through a nearest neighbor search algorithm.

8. An electronic device, comprising: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the three-dimensional point cloud line feature extraction method as described in any one of claims 1 to 5.

9. A computer-readable medium having a computer program stored thereon, wherein, When the computer program is executed by the processor, it implements the three-dimensional point cloud linear feature extraction method as described in any one of claims 1 to 5.

Images