Semantic segmentation method and device for multi-source heterogeneous point cloud, equipment and storage medium

Abstract

The invention provides a semantic segmentation method and device for multi-source heterogeneous point cloud, equipment and a storage medium. The method comprises the following steps: acquiring image data of a target area and point cloud data of each source; extracting attribute characteristics of the point cloud data of each source, and performing attribute alignment and matching on the point cloud data of each source by adopting a pre-constructed attribute alignment model to obtain a matching result of the point cloud data of each source; fusing the matching results of the point cloud data of each source to obtain fused point cloud data; and performing semantic segmentation on the fused point cloud data based on the image data to obtain a semantic segmentation result of the target region. Through the method disclosed by the invention, the semantic segmentation accuracy of the multi-source heterogeneous point cloud data can be improved.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technology, and in particular to a semantic segmentation method, device, equipment and storage medium for multi-source heterogeneous point clouds. Background Art

[0002] Point cloud semantic segmentation is to divide point cloud data into different semantic categories, such as buildings, roads, vehicles, vegetation, etc. Its function is to achieve fine semantic judgment on each point in the point cloud, so as to obtain more detailed semantic structure information and provide rich semantic features for detection, recognition, modeling and interaction.

[0003] Due to environmental restrictions, in order to obtain more scene information in an area, point cloud data is usually collected through different devices, resulting in the point cloud data of the area being multi-source heterogeneous. However, current semantic segmentation technology mainly focuses on processing point cloud data from a single source, and the semantic segmentation results of multi-source heterogeneous point cloud data are not accurate. Therefore, how to improve the accuracy of semantic segmentation of multi-source heterogeneous point cloud data is a technical problem that needs to be solved urgently by technicians in this field. Summary of the invention

[0004] In view of this, the present disclosure proposes a semantic segmentation method, apparatus, device and storage medium for multi-source heterogeneous point cloud, which can improve the semantic segmentation accuracy of multi-source heterogeneous point cloud data.

[0005] According to a first aspect of the present disclosure, a semantic segmentation method for multi-source heterogeneous point clouds is provided, comprising: Obtain image data of the target area and point cloud data from each source; Extract the attribute features of each source point cloud data, use the pre-built attribute alignment model to align and match the attributes of each source point cloud data, and obtain the matching results of each source point cloud data; The matching results of the point cloud data from each source are then fused to obtain fused point cloud data; The fused point cloud data is semantically segmented based on the image data to obtain a semantic segmentation result of the target area.

[0006] In a possible implementation, when extracting attribute features of each source point cloud data, using a pre-built attribute alignment model, performing attribute alignment and matching on each source point cloud data, and obtaining matching results of each source point cloud data, it includes: Convert the attribute features of each source point cloud data into a standard attribute feature to obtain the standard attribute features of each source point cloud data; Based on the standard attribute characteristics of the point cloud data from each source, the point cloud data from each source are matched to obtain the matching results of the point cloud data from each source.

[0007] In a possible implementation, when matching the source point cloud data based on the standard attribute features of the source point cloud data to obtain the matching results of the source point cloud data, it includes: Determine the target point cloud data from each source point cloud data; Traversing each point cloud in the target point cloud data; For the traversed current point cloud, the similarity between the current point cloud and the standard attribute features of each point cloud in other source point cloud data is calculated in turn, and the point clouds in other source point cloud data whose similarity is greater than a set threshold are matched with the current point cloud; After the traversal is completed, the matching results of the point cloud data from each source can be obtained.

[0008] In a possible implementation, when calculating the similarity between the current point cloud and the standard attribute features of each point cloud in other source point cloud data, it is implemented based on preset attribute feature weights.

[0009] In a possible implementation, the attribute feature weight is updated according to the similarity between the current point cloud and the standard attribute features of each point cloud in other source point cloud data.

[0010] In a possible implementation, when semantic segmentation is performed on the fused point cloud data based on the image data to obtain a semantic segmentation result of the target area, the method includes: Performing semantic analysis on the image data to obtain a two-dimensional semantic mask of the image data; Converting the two-dimensional semantic mask into a three-dimensional semantic mask, and aligning the three-dimensional semantic mask with the fused point cloud data; Based on the three-dimensional semantic mask aligned with the fused point cloud data, the fused point cloud data is semantically segmented to obtain a semantic segmentation result of the target area.

[0011] In a possible implementation, when performing semantic analysis on the image data to obtain a two-dimensional semantic mask of the image data, it is implemented based on a SAM2 model.

[0012] According to a second aspect of the present disclosure, a semantic segmentation method and apparatus for multi-source heterogeneous point clouds is provided, comprising: A data acquisition module is used to acquire image data of the target area and point cloud data from various sources; The attribute alignment module is used to extract the attribute features of each source point cloud data, and use the pre-built attribute alignment model to perform attribute alignment and matching on each source point cloud data to obtain the matching results of each source point cloud data; The point cloud fusion module is used to fuse the matching results of point cloud data from various sources to obtain fused point cloud data; The semantic segmentation module is used to perform semantic segmentation on the fused point cloud data based on the image data to obtain a semantic segmentation result of the target area.

[0013] According to a third aspect of the present disclosure, a semantic segmentation method and device for multi-source heterogeneous point clouds is provided, comprising: a processor; a memory for storing processor executable instructions; wherein the processor is configured to execute the method described in the first aspect of the present disclosure.

[0014] According to a fourth aspect of the present disclosure, a non-volatile computer-readable storage medium is provided, on which computer program instructions are stored, wherein the computer program instructions, when executed by a processor, implement the method described in the first aspect of the present disclosure.

[0015] The present disclosure provides a semantic segmentation method, device, equipment and storage medium for multi-source heterogeneous point clouds, the method comprising: obtaining image data of a target area and point cloud data from each source; extracting attribute features of the point cloud data from each source, using a pre-built attribute alignment model, aligning and matching the attribute of the point cloud data from each source, and obtaining matching results of the point cloud data from each source; fusing the matching results of the point cloud data from each source to obtain fused point cloud data; semantically segmenting the fused point cloud data based on the image data to obtain the semantic segmentation results of the target area. The method disclosed in the present disclosure can improve the accuracy of semantic segmentation of multi-source heterogeneous point cloud data.

[0016] Further features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate exemplary embodiments, features, and aspects of the disclosure and, together with the description, serve to explain the principles of the disclosure.

[0018] Figure 1 A flowchart showing a semantic segmentation method for multi-source heterogeneous point clouds according to an embodiment of the present disclosure is shown; Figure 2 Showing the reservoir water point cloud data according to an embodiment of the present disclosure; Figure 3 The underwater point cloud data of a reservoir according to an embodiment of the present disclosure is shown; Figure 4 The semantic segmentation results of reservoir surface point cloud data and underwater point cloud data according to an embodiment of the present disclosure are shown; Figure 5 A schematic block diagram showing a semantic segmentation method and apparatus for multi-source heterogeneous point clouds according to an embodiment of the present disclosure is shown; Figure 6 A schematic block diagram of a semantic segmentation method and device for multi-source heterogeneous point clouds according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0019] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0020] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0021] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present disclosure.

[0022] <Method Example> Figure 1 A flowchart of a semantic segmentation method for multi-source heterogeneous point clouds according to an embodiment of the present disclosure is shown. Figure 1 As shown, the method includes steps S1100-S1400.

[0023] S1100, obtaining image data of the target area and point cloud data from various sources. The target area is the geographical area to be analyzed. The image data of the target area is the remote sensing image of the target area. The point cloud data of the target area obtained by different detection equipment is the point cloud data from various sources of the target area. For example, the target area is the area where a reservoir is located. When the point cloud data of the target area is obtained, the water point cloud data of the target area is obtained by using a drone. The water point cloud data is as follows: Figure 2 As shown in the figure, the underwater point cloud data of the target area is obtained by an unmanned boat. Figure 3 As shown; the above-water point cloud data and underwater point cloud data of the target area are point cloud data from two sources of the target area.

[0024] S1200, extracting attribute features of each source point cloud data, using a pre-built attribute alignment model, performing attribute alignment and matching on each source point cloud data, and obtaining matching results of each source point cloud data.

[0025] Before executing this step, it is necessary to build an attribute alignment model. When building the attribute alignment model, it includes: building a training sample set. The training sample set includes multiple training samples, each of which includes point cloud data from different sources in the same area, as well as standard attribute features annotated for point cloud data from different sources and matching results corresponding to point cloud data from different sources. The pre-selected neural network model is trained using the training sample set. After the training is completed, the attribute alignment model can be obtained.

[0026] After obtaining the attribute alignment model, the pre-built attribute alignment model can be used to align and match the attributes of each source point cloud data. Specifically, the attribute features of each source point cloud data are extracted and input into the attribute alignment model. The attribute alignment model can automatically align and match the attributes of each source point cloud data and output the matching results of each source point cloud data.

[0027] When the attribute alignment model performs attribute alignment and matching on each source point cloud data and outputs the matching result of each source point cloud data, the following steps may be included: First, the attribute features of the point cloud data from each source are converted into standard attribute features to obtain the standard attribute features of the point cloud data from each source. The standard attribute features are pre-set and used to uniformly represent the attribute features of point cloud data from different sources, thereby avoiding point cloud matching errors caused by inconsistent attribute feature representations between different data sources.

[0028] Second, based on the standard attribute features of each source point cloud data, each source point cloud data is matched to obtain the matching results of each source point cloud data. The specific matching process may include the following steps: First, the target point cloud data is determined from each source point cloud data. Specifically, point cloud data from one source can be randomly selected from point cloud data from multiple sources as the target point cloud data. After the target point cloud data is selected, the point cloud data from other sources are matched with the target point cloud data respectively to obtain the matching results of the point cloud data from each source.

[0029] Secondly, traverse each point cloud in the target point cloud data, and for the current point cloud traversed, calculate the similarity between the current point cloud and the standard attribute features of each point cloud in other source point cloud data in turn, and match the point clouds in other source point cloud data with similarities greater than the set threshold with the current point cloud.

[0030] Finally, after the traversal is completed, the matching results of the point cloud data from each source can be obtained.

[0031] In a possible implementation, when calculating the similarity between the standard attribute features of the current point cloud and each point cloud in other source point cloud data, it is implemented based on a preset attribute feature weight. Specifically, a point cloud data usually includes multiple standard attribute features, and a corresponding attribute feature weight is pre-set for each attribute feature. In this way, when calculating the similarity between the standard attribute features of the current point cloud and each point cloud in other source point cloud data, the similarity between the same standard attribute features will be calculated respectively, and then the calculated similarities will be weighted and summed by the attribute feature weights corresponding to each standard attribute feature, so as to obtain the similarity between the standard attribute features of the current point cloud and each point cloud in the other source point cloud data.

[0032] For example, the standard attribute features are position, color and normal vector. The attribute feature weight set for the standard attribute feature of position is W1, the attribute feature weight set for the standard attribute feature of color is W2, and the attribute feature weight set for the standard attribute feature of normal vector is W3. P1 is the current point cloud in the target point cloud data, Q1 is a point cloud in other source point cloud data, and the standard attribute features corresponding to P1 and Q1 are shown in Table 1.

[0033] Table 1 In this example, when calculating the similarity between the standard attribute features of P1 and Q1, the similarity S1 between position a and position b is calculated first, and then the similarity S2 between color b and color c is calculated, and then the similarity between normal vector c and normal vector f is calculated. Finally, the attribute feature weight W1 corresponding to the position, the attribute feature weight W2 corresponding to the color, and the attribute feature weight W3 corresponding to the normal vector are used to weighted sum the similarity S1 between the positions, the similarity S2 between the colors, and the similarity S3 between the normal vectors, respectively, to obtain the similarity between the standard attribute features of P1 and Q1, that is, the calculation formula for the similarity S between the standard attribute features of P1 and Q1 is as follows: S=W1×S1+W2×S2+W3×S3 By referring to the above method, the similarity between the current point cloud and the standard attribute features of each point cloud in other source point cloud data can be calculated. After calculating the similarity between the standard attribute features of each point cloud in the current point cloud and other source point cloud data, the point clouds in other source point cloud data with similarity greater than the set threshold are matched with the current point cloud. In this way, after the traversal is completed, the matching results of each source point cloud data can be obtained.

[0034] In one possible implementation, in order to improve the accuracy of matching between point cloud data, the preset attribute feature weights are updated based on the similarity between the current point cloud and the standard attribute features of each point cloud in other source point cloud data. Continuing with the above embodiment, after calculating the similarity S between the standard attribute features of P1 and Q1, the similarity S1 between positions is used to update the attribute feature weight W1 corresponding to the position, the similarity S2 between colors is used to update the attribute feature weight W2 corresponding to the color, and the similarity S3 between normal vectors is used to update the attribute feature weight W3 corresponding to the normal vector, so that the similarity between the standard attribute features between subsequent point clouds can be calculated based on the updated weights. Among them, the update formulas for each attribute feature weight are as follows: Update weight of attribute feature weight W1 = S1×W1; Update weight of attribute feature weight W2 = S2×W2; The updated weight of the attribute feature weight W3 = S3×W3.

[0035] In a possible implementation, in order to improve the accuracy of the matching results of the point cloud data from each source, the semantic information of each point cloud in the point cloud data from each source will also be calculated, and the semantic information of each point cloud in the point cloud data from each source will be used as a feature attribute information of each point cloud data, so as to perform attribute alignment and matching of the point cloud data from each source based on the feature attributes including the semantic information. When calculating the semantic information of each point cloud in the point cloud data from each source, the above-mentioned image data can be first calculated by SAM2 for semantic segmentation to obtain a two-dimensional semantic mask, and then the two-dimensional semantic mask is converted into a three-dimensional semantic mask. Then, the three-dimensional semantic mask is aligned and matched with the point cloud data from each source, and the semantic information in the three-dimensional semantic mask is assigned as a feature attribute to the point cloud in the point cloud data from each source that matches the three-dimensional semantic mask. In this way, the point cloud data from each source has the feature attribute of semantic information.

[0036] After obtaining the matching results of the point cloud data from each source, step S1300 can be executed to further fuse the matching results of the point cloud data from each source to obtain fused point cloud data. Specifically, the matching results will include multiple groups of matching point clouds. For each group of matching point clouds, each group of matching point clouds will be fused into one point cloud, and the values ​​of each standard attribute feature of the fused point cloud will be equal to the weighted average value between the same standard attribute features of multiple matching point clouds. Continuing with the above embodiment, when P1 and Q1 are matching point clouds, the position of the fused point cloud is equal to the weighted average value of position a and position b, the color is equal to the weighted average value of color b and color c, and the normal vector is equal to the weighted average value of normal vector c and normal vector f. After obtaining all the fused point cloud data, the fused point cloud data can be inserted into the point cloud set composed of the point cloud data from each source by interpolation or fitting to obtain the fused point cloud data.

[0037] After the fused point cloud data is obtained, step S1400 can be executed to perform semantic segmentation on the fused point cloud data based on the image data to obtain the semantic segmentation result of the target area. The specific steps can be as follows: First, semantic analysis is performed on the image data to obtain a two-dimensional semantic mask of the image data. Specifically, the image data can be input into a SAM2 model to output a two-dimensional semantic mask of the image data through the SAM2 model.

[0038] Second, the two-dimensional semantic mask is converted into a three-dimensional semantic mask, and the three-dimensional semantic mask is aligned with the fused point cloud data. Specifically, the camera parameters corresponding to the image data are first obtained, wherein the camera parameters include the intrinsic and extrinsic information of the camera, such as the focal length, principal point coordinates, distortion parameters, etc. of the camera. Then, the two-dimensional semantic mask is converted into a three-dimensional semantic mask using the camera parameters. Specifically, the two-dimensional semantic mask includes multiple pixels, each pixel has a two-dimensional pixel coordinate and semantics, and the semantics is used to characterize the category or boundary features of the pixel. For each pixel, the two-dimensional pixel coordinates are converted into the corresponding three-dimensional space coordinates using the camera parameters, and the converted three-dimensional space coordinates are bound to the corresponding semantics. After all pixel processing is completed, the corresponding three-dimensional semantic mask can be obtained. That is, the three-dimensional semantic mask includes multiple spatial points, and semantic information is annotated for each spatial point. After obtaining the three-dimensional semantic mask, the point cloud data in the three-dimensional semantic mask is subjected to preprocessing operations such as noise removal and smoothing, so as to output a more accurate and complete three-dimensional semantic mask. Finally, the preprocessed 3D semantic mask and the fused point cloud data are aligned and matched with each other in attributes. The specific attribute alignment and matching process is shown in step S1200 and will not be described in detail here.

[0039] Third, based on the three-dimensional semantic mask aligned with the fused point cloud data, the fused point cloud data is semantically segmented to obtain the semantic segmentation result of the target area. Specifically, the following steps may be included: First, the three-dimensional semantic mask is fused with the matching point cloud in the fused point cloud data. The specific fusion method is described above and will not be repeated here. Then, for the fused point cloud data, the bidirectional group overlap algorithm is used to merge adjacent point clouds to reduce overlaps and gaps and improve the integrity and accuracy of the point cloud data. Finally, the regional merging method is used to further extract and merge the semantic information of different regions, and finally obtain a complete 3D semantic segmentation result.

[0040] In the embodiment where the target area is a reservoir, after fusing the above-water point cloud data and the underwater point cloud data, the following is obtained: Figure 4 The semantic segmentation results shown. Figure 4 It can be seen that the semantic segmentation method disclosed in the present invention can not only well display the above-water point cloud features, but also significantly display the lake surface features in the underwater point cloud data, so that the semantic segmentation results can show richer scene information.

[0041] The present disclosure provides a semantic segmentation method for multi-source heterogeneous point clouds, including: obtaining image data of a target area and point cloud data from each source; extracting attribute features of the point cloud data from each source, using a pre-built attribute alignment model to perform attribute alignment and matching on the point cloud data from each source, and obtaining matching results of the point cloud data from each source; fusing the matching results of the point cloud data from each source to obtain fused point cloud data; and performing semantic segmentation on the fused point cloud data based on the image data to obtain a semantic segmentation result of the target area. The method disclosed in the present disclosure can improve the accuracy of semantic segmentation of multi-source heterogeneous point cloud data.

[0042] <Device Example> Figure 5 FIG. 1 is a schematic block diagram of a semantic segmentation method and apparatus for multi-source heterogeneous point clouds according to an embodiment of the present disclosure. Figure 5 As shown, the device 100 includes: The data acquisition module 110 is used to acquire the image data of the target area and the point cloud data of each source; The attribute alignment module 120 is used to extract attribute features of each source point cloud data, and use a pre-built attribute alignment model to perform attribute alignment and matching on each source point cloud data to obtain matching results of each source point cloud data; A point cloud fusion module 130 is used to fuse the matching results of the point cloud data from various sources to obtain fused point cloud data; The semantic segmentation module 140 is used to perform semantic segmentation on the fused point cloud data based on the image data to obtain a semantic segmentation result of the target area.

[0043] <Equipment Embodiment> Figure 6 A schematic block diagram of a semantic segmentation method and device for multi-source heterogeneous point clouds according to an embodiment of the present disclosure is shown. Figure 6 As shown, a semantic segmentation method and device 200 for multi-source heterogeneous point clouds includes: a processor 210 and a memory 220 for storing executable instructions of the processor 210. The processor 210 is configured to implement any of the above-mentioned semantic segmentation methods for multi-source heterogeneous point clouds when executing the executable instructions.

[0044] Here, it should be noted that the number of processors 210 may be one or more. At the same time, in the semantic segmentation method and device 200 of multi-source heterogeneous point cloud in the embodiment of the present disclosure, an input device 230 and an output device 240 may also be included. Among them, the processor 210, the memory 220, the input device 230 and the output device 240 may be connected through a bus or in other ways, which are not specifically limited here.

[0045] The memory 220 is a computer-readable storage medium that can be used to store software programs, computer executable programs, and various modules, such as the program or module corresponding to the semantic segmentation method of multi-source heterogeneous point clouds in the embodiment of the present disclosure. The processor 210 executes various functional applications and data processing of the semantic segmentation method of multi-source heterogeneous point clouds device 200 by running the software programs or modules stored in the memory 220.

[0046] The input device 230 may be used to receive input numbers or signals. The signals may be key signals related to user settings and function control of the device / terminal / server. The output device 240 may include display devices such as display screens.

[0047] <Storage Medium Embodiment> According to a fourth aspect of the present disclosure, a non-volatile computer-readable storage medium is also provided, on which computer program instructions are stored. When the computer program instructions are executed by the processor 210, any of the semantic segmentation methods for multi-source heterogeneous point clouds described above is implemented.

[0048] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A semantic segmentation method for multi-source heterogeneous point clouds, characterized in that: include: Obtain image data of the target area and point cloud data from various sources; Extract the attribute features of each source point cloud data, use the pre-built attribute alignment model to align and match the attributes of each source point cloud data, and obtain the matching results of each source point cloud data; The matching results of point cloud data from each source are then fused to obtain fused point cloud data; The fused point cloud data is semantically segmented based on the image data to obtain a semantic segmentation result of the target area.

2. The method according to claim 1, characterized in that When extracting attribute features of each source point cloud data, using a pre-built attribute alignment model to perform attribute alignment and matching on each source point cloud data, and obtaining matching results for each source point cloud data, the following steps are included: Converting the attribute features of each source point cloud data into a standard attribute feature to obtain the standard attribute features of each source point cloud data; Based on the standard attribute characteristics of the point cloud data from each source, the point cloud data from each source are matched to obtain the matching results of the point cloud data from each source.

3. The method according to claim 2, characterized in that When matching the point cloud data from each source based on the standard attribute characteristics of the point cloud data from each source and obtaining the matching results of the point cloud data from each source, the following steps are included: Determine target point cloud data from each source point cloud data; Traversing each point cloud in the target point cloud data; For the traversed current point cloud, the similarity between the current point cloud and the standard attribute features of each point cloud in other source point cloud data is calculated in sequence, and the point clouds in other source point cloud data with similarity greater than a set threshold are considered to be matched with the current point cloud; After the traversal is completed, the matching results of the point cloud data from each source can be obtained.

4. The method according to claim 3, characterized in that When calculating the similarity between the current point cloud and the standard attribute features of each point cloud in other source point cloud data, it is achieved based on the preset attribute feature weights.

5. The method according to claim 4, characterized in that The attribute feature weight is updated according to the similarity between the current point cloud and the standard attribute features of each point cloud in other source point cloud data.

6. The method according to claim 1, characterized in that When semantic segmentation is performed on the fused point cloud data based on the image data to obtain a semantic segmentation result of the target area, the method includes: Performing semantic analysis on the image data to obtain a two-dimensional semantic mask of the image data; Converting the two-dimensional semantic mask into a three-dimensional semantic mask, and aligning the three-dimensional semantic mask with the fused point cloud data; Based on the three-dimensional semantic mask aligned with the fused point cloud data, semantic segmentation is performed on the fused point cloud data to obtain a semantic segmentation result of the target area.

7. The method according to claim 6, characterized in that When semantic analysis is performed on the image data to obtain a two-dimensional semantic mask of the image data, it is implemented based on the SAM2 model.

8. A semantic segmentation device for multi-source heterogeneous point clouds, characterized in that: include: Data acquisition module, used to obtain image data of the target area and point cloud data from various sources; The attribute alignment module is used to extract the attribute features of each source point cloud data, use the pre-built attribute alignment model to perform attribute alignment and matching on each source point cloud data, and obtain the matching results of each source point cloud data; The point cloud fusion module is used to fuse the matching results of point cloud data from various sources to obtain fused point cloud data; The semantic segmentation module is used to perform semantic segmentation on the fused point cloud data based on the image data to obtain a semantic segmentation result of the target area.

9. A semantic segmentation device for multi-source heterogeneous point clouds, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to implement the method according to any one of claims 1 to 7 when executing the executable instructions.

10. A non-volatile computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images