CN115471513B - Point cloud segmentation method and device - Google Patents
Point cloud segmentation method and device Download PDFInfo
- Publication number
- CN115471513B CN115471513B CN202211357371.6A CN202211357371A CN115471513B CN 115471513 B CN115471513 B CN 115471513B CN 202211357371 A CN202211357371 A CN 202211357371A CN 115471513 B CN115471513 B CN 115471513B
- Authority
- CN
- China
- Prior art keywords
- point cloud
- point
- voxel
- cloud data
- feature
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000011218 segmentation Effects 0.000 title claims abstract description 77
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000000605 extraction Methods 0.000 claims abstract description 43
- 230000002093 peripheral effect Effects 0.000 claims description 47
- 238000012545 processing Methods 0.000 claims description 28
- 238000006243 chemical reaction Methods 0.000 claims description 14
- 238000004364 calculation method Methods 0.000 abstract description 17
- 230000000875 corresponding effect Effects 0.000 description 17
- 230000008569 process Effects 0.000 description 12
- 238000010586 diagram Methods 0.000 description 7
- 238000002310 reflectometry Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 5
- 238000004590 computer program Methods 0.000 description 5
- 238000012935 Averaging Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000002485 combustion reaction Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 239000000446 fuel Substances 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4038—Image mosaicing, e.g. composing plane images from plane sub-images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2200/00—Indexing scheme for image data processing or generation, in general
- G06T2200/32—Indexing scheme for image data processing or generation, in general involving image mosaicing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30248—Vehicle exterior or interior
- G06T2207/30252—Vehicle exterior; Vicinity of vehicle
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The disclosure relates to a point cloud segmentation method and device, wherein the method comprises the following steps: acquiring target point cloud data, wherein the target point cloud data is obtained by scanning target objects around a vehicle by laser beams; performing cylinder voxel division and feature extraction on the target point cloud data to obtain first point cloud features of the target point cloud data; performing point cloud feature extraction on the target point cloud data to acquire a second point cloud feature of the target point cloud data; according to the first point Yun Te and the second point cloud feature, the point cloud segmentation result of the target point cloud data is determined, so that the first point Yun Te determined based on voxel division and the second point cloud feature obtained by point cloud feature extraction can be combined, the point cloud segmentation result is determined, the requirement on the voxel size is not high, the voxel size of the point cloud is larger at the position where the point cloud is more sparse, under the condition that the point cloud segmentation accuracy is ensured, the calculation amount is greatly reduced, and the calculation time is shortened.
Description
Technical Field
The disclosure relates to the technical field of automatic driving, in particular to a point cloud segmentation method and device.
Background
Currently, in the field of automatic driving, a vehicle needs to perform three-dimensional reconstruction on a peripheral area, determine three-dimensional information, a position and the like of an object in the peripheral area, and further avoid an obstacle object or guide the vehicle to run. Before three-dimensional reconstruction, peripheral areas need to be scanned to obtain point cloud data, the point cloud data are divided to divide the point cloud data corresponding to different types of objects, so that point cloud modeling is performed on each object respectively, and three-dimensional information, positions and the like of the objects are determined.
The point cloud segmentation algorithm in the related art is, for example, a point-based method or a voxel-based method. In the point-based method, due to the irregularity of point cloud data, the calculation amount for searching the adjacent points is large, and the time consumption is long. In the voxel-based method, the size of the voxel is very small, the calculation amount is large, and the time is long if a high-resolution segmentation result is to be obtained.
Disclosure of Invention
The disclosure provides a point cloud segmentation method and device.
According to a first aspect of an embodiment of the present disclosure, there is provided a point cloud segmentation method, including: acquiring target point cloud data, wherein the target point cloud data is obtained by scanning target objects around a vehicle by laser beams; performing cylinder voxel division and feature extraction on the target point cloud data to acquire a first point cloud feature of the target point cloud data; performing point cloud feature extraction on the target point cloud data to acquire a second point cloud feature of the target point cloud data; and determining a point cloud segmentation result of the target point cloud data according to the first point Yun Te and the second point cloud characteristic.
In an embodiment of the present disclosure, the performing cylinder voxel division and feature extraction on the target point cloud data to obtain a first point cloud feature of the target point cloud data includes: performing coordinate system conversion on the target point cloud data to obtain first point cloud data under a cylindrical coordinate system; carrying out voxel division on the first point cloud data to obtain a plurality of voxels; the voxel is a fan-ring cylinder voxel, and the upper bottom surface and the lower bottom surface of the fan-ring cylinder voxel are in a fan ring shape; determining voxel characteristics of a plurality of voxels according to local point cloud data in the plurality of voxels; and determining the first point cloud characteristic according to the voxel characteristics of the voxels and the first point cloud data.
In one embodiment of the present disclosure, the determining voxel characteristics of a plurality of the voxels from local point cloud data within the plurality of voxels comprises: for each voxel, determining voxel data of the voxel according to local point cloud data, located in the voxel, in the first point cloud data; and performing voxel characteristic extraction processing on the voxel data of the voxels to obtain the voxel characteristics of the voxels.
In an embodiment of the present disclosure, the determining the first point cloud feature according to the voxel features of the plurality of voxels and the first point cloud data includes: determining center point coordinate information of a plurality of voxels; for each point in the first point cloud data, determining a point feature of the point according to center point coordinate information of a plurality of peripheral voxels of the point, voxel features of the plurality of peripheral voxels, and coordinate information of the point in the first point cloud data; and determining the first point cloud characteristics according to the point characteristics of the points and the point data of the points in the target point cloud data.
In one embodiment of the present disclosure, the determining, for each point in the first point cloud data, a point feature of the point according to center point coordinate information of a plurality of peripheral voxels of the point, voxel features of the plurality of peripheral voxels, and coordinate information of the point in the first point cloud data includes: for each point in the first point cloud data, determining the weight of the voxel characteristics of a plurality of peripheral voxels according to the central point coordinate information of a plurality of peripheral voxels of the point and the coordinate information of the point in the first point cloud data; and determining the point characteristics of the points according to the voxel characteristics of the peripheral voxels and the corresponding weights.
In an embodiment of the disclosure, the determining a point cloud segmentation result of the target point cloud data according to the first point Yun Te and the second point cloud feature includes: splicing the first point Yun Te and the second point cloud feature to obtain a spliced point cloud feature; determining the weight of each dimension of the point feature in the spliced point cloud feature; according to the weight of each dimension, carrying out weighting processing on the corresponding dimension feature of the point feature in the spliced point cloud feature to obtain a processed point cloud feature; classifying the processed point cloud characteristics to obtain a classification result of each point in the target point cloud data; and determining the point cloud segmentation result according to the classification result of each point in the target point cloud data.
According to a second aspect of the embodiments of the present disclosure, there is also provided a point cloud segmentation apparatus, the apparatus including: the system comprises an acquisition module, a processing module and a display module, wherein the acquisition module is used for acquiring target point cloud data, and the target point cloud data is obtained by scanning a target object around a vehicle by a laser beam; the first processing module is used for performing cylinder voxel division and feature extraction on the target point cloud data to acquire a first point cloud feature of the target point cloud data; the second processing module is used for extracting point cloud characteristics of the target point cloud data to obtain second point cloud characteristics of the target point cloud data; and the determining module is used for determining a point cloud segmentation result of the target point cloud data according to the first point Yun Te and the second point cloud characteristic.
In one embodiment of the present disclosure, the first processing module includes: the device comprises a conversion unit, a dividing unit, a first determining unit and a second determining unit; the conversion unit is used for carrying out coordinate system conversion on the target point cloud data to obtain first point cloud data under a cylindrical coordinate system; the dividing unit is used for carrying out voxel division on the first point cloud data to obtain a plurality of voxels; the voxel is a fan-ring cylinder voxel, and the upper bottom surface and the lower bottom surface of the fan-ring cylinder voxel are in a fan-ring shape; the first determining unit is used for determining voxel characteristics of a plurality of voxels according to local point cloud data in the voxels; the second determining unit is configured to determine the first point cloud feature according to voxel features of the voxels and the first point cloud data.
In an embodiment of the present disclosure, the first determining unit is specifically configured to, for each voxel, determine voxel data of the voxel according to local point cloud data located within the voxel in the first point cloud data; and performing voxel characteristic extraction processing on the voxel data of the voxels to obtain the voxel characteristics of the voxels.
In an embodiment of the present disclosure, the second determining unit is specifically configured to determine center point coordinate information of a plurality of voxels; for each point in the first point cloud data, determining a point feature of the point according to center point coordinate information of a plurality of peripheral voxels of the point, voxel features of the plurality of peripheral voxels, and coordinate information of the point in the first point cloud data; and determining the first point cloud characteristics according to the point characteristics of the points and the point data of the points in the target point cloud data.
In an embodiment of the present disclosure, the second determining unit is specifically configured to, for each point in the first point cloud data, determine a weight of a voxel characteristic of a plurality of peripheral voxels according to center point coordinate information of a plurality of peripheral voxels of the point and coordinate information of the point in the first point cloud data; and determining the point characteristics of the points according to the voxel characteristics of the peripheral voxels and the corresponding weights.
In an embodiment of the disclosure, the determining module is specifically configured to perform a stitching process on the first point Yun Te and the second point cloud feature to obtain a stitched point cloud feature; determining the weight of each dimension of the point feature in the spliced point cloud feature; according to the weight of each dimension, carrying out weighting processing on corresponding dimension features of the point feature in the spliced point cloud feature to obtain processed point cloud features; classifying the processed point cloud characteristics to obtain a classification result of each point in the target point cloud data; and determining the point cloud segmentation result according to the classification result of each point in the target point cloud data.
According to a third aspect of the embodiments of the present disclosure, there is also provided a vehicle including: a processor;
a memory for storing processor-executable instructions; wherein the processor is configured to: the steps of the point cloud segmentation method as described above are implemented.
According to a fourth aspect of the embodiments of the present disclosure, there is also provided an electronic apparatus, including: a processor; a memory for storing the processor-executable instructions; wherein the processor is configured to execute the instructions to implement the point cloud segmentation method as described above.
According to a fifth aspect of embodiments of the present disclosure, there is also provided a non-transitory computer-readable storage medium, wherein instructions of the storage medium, when executed by a processor, enable the processor to perform the point cloud segmentation method as described above.
The technical scheme provided by the embodiment of the disclosure at least brings the following beneficial effects:
acquiring target point cloud data, wherein the target point cloud data is obtained by scanning target objects around a vehicle by laser beams; performing cylinder voxel division and feature extraction on the target point cloud data to acquire first point cloud features of the target point cloud data; performing point cloud feature extraction on the target point cloud data to acquire a second point cloud feature of the target point cloud data; according to the first point Yun Te and the second point cloud feature, the point cloud segmentation result of the target point cloud data is determined, so that the first point Yun Te determined based on voxel division and the second point cloud feature obtained by point cloud feature extraction can be combined, the point cloud segmentation result is determined, the requirement on the voxel size is not high, the voxel size of the point cloud is larger at the position where the point cloud is more sparse, under the condition that the point cloud segmentation accuracy is ensured, the calculation amount is greatly reduced, and the calculation time is shortened.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the disclosure and are not to be construed as limiting the disclosure.
FIG. 1 is a flow chart of a point cloud segmentation method according to an embodiment of the present disclosure;
FIG. 2 is a flow chart of a point cloud segmentation method according to another embodiment of the present disclosure;
FIG. 3 is a flow chart of a point cloud segmentation method according to another embodiment of the present disclosure;
FIG. 4 is a schematic structural diagram of a point cloud segmentation apparatus according to an embodiment of the disclosure;
FIG. 5 is a block diagram illustrating an electronic device according to an exemplary embodiment of the present disclosure;
FIG. 6 is a functional block diagram schematic of a vehicle shown in an exemplary embodiment of the present disclosure.
Detailed Description
In order to make the technical solutions of the present disclosure better understood by those of ordinary skill in the art, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.
It should be noted that the terms "first," "second," and the like in the description and claims of the present disclosure and in the above-described drawings are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the disclosure described herein are capable of operation in sequences other than those illustrated or otherwise described herein. The implementations described in the exemplary embodiments below do not represent all implementations consistent with the present disclosure.
Fig. 1 is a flowchart of a point cloud segmentation method according to an embodiment of the present disclosure. It should be noted that the point cloud segmentation method of this embodiment is executed by a point cloud segmentation apparatus, where the point cloud segmentation apparatus may be implemented by software and/or hardware, and the point cloud segmentation apparatus may be configured in an electronic device or a vehicle, and the electronic device may include a mobile terminal (e.g., a mobile phone), a terminal device such as a tablet computer, or a server. The following description will be given taking an execution body as an electronic device as an example.
As shown in fig. 1, the method comprises the steps of:
In the embodiment of the disclosure, the laser beam can be used for scanning the surrounding target objects in real time during the driving process of the vehicle. The data obtained by scanning each time point is a frame of point cloud data, and then multi-frame point cloud data is obtained. The target point cloud data in the present application refers to one frame of point cloud data. Such as, among others, the ground, buildings, trees, vehicles, signboards, etc.
In the disclosed embodiment, the target point cloud data may include point data of a plurality of points, wherein the point data may include at least one of the following parameters: coordinate information, reflectivity, serial number of laser line, and the number of points scanned by the laser line. The coordinate information is coordinate information of a point under the vehicle body coordinate system. The vehicle body coordinate system may use an intersection of a rear axis center line of the vehicle and the ground as an origin, a direction toward the right side of the vehicle as an X-axis forward direction, a direction toward the front side of the vehicle as a Y-axis forward direction, and a direction toward the upper side of the vehicle as a Z-axis forward direction. Thus, the coordinate information may include X-axis coordinates, Y-axis coordinates, and Z-axis coordinates relative to the origin of coordinates.
The reflectivity is the reflectivity of the laser line on the surface of the object after the laser line is projected on the surface of the object. The serial number of the laser line refers to the number of the laser line in the laser line that scanned the spot. The laser beam may include, for example, 128 lines, which are numbered 0 to 127.
In the embodiment of the disclosure, coordinate system conversion can be performed on target point cloud data in a vehicle body coordinate system to obtain first point cloud data in a cylindrical coordinate system, and then voxel division, voxel feature extraction and point feature extraction are performed to obtain first point cloud features.
The cylindrical coordinate system can use the intersection point of the central line of the rear axle of the vehicle and the ground as an origin, and is provided with a reference line using the circle center as a starting point, the reference line extends along the positive and negative directions of the Z axis to obtain a reference surface, and the reference surface passes through the axis of the cylindrical coordinate system; and taking the distance from the point to the origin as a radius, taking the included angle between a connecting line between the point and the origin and the reference surface as an angle, and taking the distance from the projection of the point on the Z axis to the origin as a Z value. Therefore, the coordinate information in the cylindrical coordinate system may include: radius from the origin, angle, and Z-axis coordinate.
In the embodiment of the present disclosure, voxel division is performed on the first point cloud data in the cylindrical coordinate system, the obtained voxel is a sector-ring cylinder voxel, and the upper bottom surface and the lower bottom surface of the sector-ring cylinder voxel are in a sector ring shape. The hollow ring is sheared according to one or more straight lines passing through the center of the circle, and the obtained local area is a sector ring. The multiple sector-ring cylinder voxels obtained through division under the cylindrical coordinate system have the same axis, namely the axis of the cylindrical coordinate system.
In the target point cloud data, based on the principle that a laser beam scans target objects around a vehicle, the farther the distance from the origin of a vehicle body coordinate system is, the more sparse the point cloud is; the closer the distance to the origin of the vehicle body coordinate system, the denser the point cloud. In the cylindrical coordinate system, the farther the distance from the origin is, the larger the size of the fan-ring cylinder voxels is; the closer the distance from the original point is, the smaller the size of the sector ring cylinder body voxel is, so that the point cloud characteristics can be extracted from the sector ring cylinder body voxel with the smaller dense point cloud combination size, and the point cloud characteristics can be extracted from the sector ring cylinder body voxel with the larger sparse point cloud combination size, so that the accuracy of characteristic extraction can be improved, the calculation amount is greatly reduced, the calculation time is shortened, and the calculation accuracy is improved.
In the embodiment of the present disclosure, the target point cloud data may be input into a Multilayer Perceptron (MLP) to obtain a second point cloud feature output by the MLP. The MLP is used for carrying out feature extraction processing on point data of a plurality of points in the target point cloud data to obtain second point cloud features. The second point cloud feature is formed by combining point data of a plurality of points and point features. For example, the number of dimensions of a point feature may be 256 dimensions, etc.; the number of points may be 30 ten thousand, etc.
And 104, determining a point cloud segmentation result of the target point cloud data according to the first point Yun Te and the second point cloud characteristics.
In the embodiment of the present disclosure, the electronic device may perform a stitching process on the first point Yun Te and the second point cloud feature to obtain a stitched point cloud feature; and determining the categories of a plurality of points in the target point cloud data according to the characteristics of the spliced point cloud, and further determining a point cloud segmentation result. The categories, such as ground, buildings, trees, vehicles, signboards, etc., may be set according to actual needs, and are not limited specifically here.
According to the point cloud segmentation method, target point cloud data are obtained, wherein the target point cloud data are obtained by scanning target objects around a vehicle through laser beams; performing cylinder voxel division and feature extraction on the target point cloud data to obtain first point cloud features of the target point cloud data; performing point cloud feature extraction on the target point cloud data to acquire a second point cloud feature of the target point cloud data; according to the first point Yun Te and the second point cloud feature, the point cloud segmentation result of the target point cloud data is determined, so that the first point Yun Te determined based on voxel division and the second point cloud feature obtained by point cloud feature extraction can be combined, the point cloud segmentation result is determined, the requirement on the voxel size is not high, the voxel size of the point cloud is larger at the position where the point cloud is more sparse, under the condition that the point cloud segmentation accuracy is ensured, the calculation amount is greatly reduced, and the calculation time is shortened.
Fig. 2 is a flowchart of a point cloud segmentation method according to another embodiment of the disclosure. As shown in fig. 2, the method comprises the steps of:
In the embodiment of the disclosure, the target point cloud data is point cloud data in a vehicle body coordinate system, and the electronic device may perform conversion processing on coordinate information of a plurality of points in the target point cloud data based on a coordinate conversion matrix from the vehicle body coordinate system to a cylindrical coordinate system to obtain coordinate information of the plurality of points in the cylindrical coordinate system; and determining first point cloud data in the cylindrical coordinate system by combining the coordinate information of the plurality of points in the cylindrical coordinate system and other point data of the plurality of points.
The number of the points in the first point cloud data is consistent with that of the points in the target point cloud data, and the points in the first point cloud data correspond to the points in the target point cloud data one to one.
In the embodiment of the present disclosure, the coordinate information in the cylindrical coordinate system may include: radius from the origin, angle, and Z-axis coordinate. The voxels in the cylindrical coordinate system can be represented by a radius range, an angle range and a Z-axis coordinate range. That is, the points where the corresponding radius is within the radius range, the corresponding angle is within the angle range, and the corresponding Z-axis coordinate is within the Z-axis coordinate range are all points within the voxel.
And step 204, determining voxel characteristics of a plurality of voxels according to the local point cloud data in the plurality of voxels.
In the embodiment of the present disclosure, the electronic device may perform the process of step 204, for example, to determine, for each voxel, voxel data of the voxel according to local point cloud data located in the voxel in the first point cloud data; and carrying out voxel characteristic extraction processing on the voxel data of the plurality of voxels to obtain the voxel characteristics of the plurality of voxels.
Taking the voxel as an example represented by a first radius range, a first angle range and a first Z-axis coordinate range, points in the local point cloud data in the voxel satisfy all the following conditions: the radius in the coordinate information of the point is located within the first radius range, the angle in the coordinate information of the point is located within the first angle range, and the Z-axis coordinate in the coordinate information of the point is located within the first Z-axis coordinate range.
For each voxel, the process of determining the voxel data of the voxel by the electronic device according to the local point cloud data located in the voxel in the first point cloud data may be, for example, summing and averaging the point data of all points in the local point cloud data to obtain the voxel data of the voxel. Taking the example that the point data of the points comprises coordinate information and reflectivity, adding the coordinate information of all the points in the local point cloud data and averaging to obtain the coordinate information of the voxel; adding and averaging the reflectivity of all points in the local point cloud data to obtain the reflectivity of the voxel; the combination of the coordinate information and the reflectivity of the voxel is the voxel data of the voxel.
The process of extracting the voxel characteristics of the multiple voxels from the voxel data of the multiple voxels by the electronic device to obtain the voxel characteristics of the multiple voxels may be, for example, inputting the voxel data of the multiple voxels into a preset 3D sparse convolution network to obtain the voxel characteristics of the multiple voxels output by the 3D sparse convolution network.
In the embodiment of the present disclosure, the electronic device executing the process of step 205 may be, for example, determining center point coordinate information of a plurality of voxels; for each point in the first point cloud data, determining a point feature of the point according to center point coordinate information of a plurality of peripheral voxels of the point, voxel features of the plurality of peripheral voxels, and coordinate information of the point in the first point cloud data; and determining the first point cloud characteristics according to the point characteristics of the points and the point data of the points in the target point cloud data.
Wherein, for each point in the first point cloud data, the electronic device may determine, based on a trilinear interpolation strategy, a point feature of the point according to center point coordinate information of a plurality of peripheral voxels of the point, voxel features of the plurality of peripheral voxels, and coordinate information of the point in the first point cloud data. The principle of determining the point characteristics of the point in combination with the trilinear interpolation strategy may be, for example, determining, for each point in the first point cloud data, weights of voxel characteristics of a plurality of peripheral voxels according to center point coordinate information of a plurality of peripheral voxels of the point and coordinate information of the point in the first point cloud data; and determining the point characteristics of the points according to the voxel characteristics of the peripheral voxels and the corresponding weights.
The process of determining the point feature of the point according to the voxel features of the plurality of peripheral voxels and the corresponding weights may be, for example, performing weighted summation on the voxel features of the plurality of peripheral voxels according to the corresponding weights to obtain the point feature of the point. The weight of the voxel characteristics of the peripheral voxels can be negatively correlated with the distance of the peripheral voxels from the point, and the larger the distance is, the smaller the weight is; the smaller the distance, the greater the weight.
The process of determining the point characteristics of the points can determine different point characteristics aiming at different points in the same voxel, so that the accuracy of point characteristic extraction is improved, the accuracy of the first point cloud characteristics is improved, and the accuracy of point cloud segmentation is improved.
And step 206, performing point cloud feature extraction on the target point cloud data to acquire a second point cloud feature of the target point cloud data.
And step 207, determining a point cloud segmentation result of the target point cloud data according to the first point Yun Te and the second point cloud characteristics.
It should be noted that, the detailed description of step 201, step 206, and step 207 may refer to the corresponding description of step 101, step 103, and step 104 in the embodiment shown in fig. 1, and a repeated description is not repeated here.
According to the point cloud segmentation method, target point cloud data are obtained, wherein the target point cloud data are obtained by scanning target objects around a vehicle through laser beams; converting a coordinate system of the target point cloud data to obtain first point cloud data under a cylindrical coordinate system; carrying out voxel division on the first point cloud data to obtain a plurality of voxels; the voxel is a fan-ring cylinder voxel, and the upper bottom surface and the lower bottom surface of the fan-ring cylinder voxel are in a fan-ring shape; determining voxel characteristics of a plurality of voxels according to local point cloud data in the plurality of voxels; determining a first point cloud characteristic according to voxel characteristics of a plurality of voxels and first point cloud data; performing point cloud feature extraction on the target point cloud data to acquire a second point cloud feature of the target point cloud data; according to the first point Yun Te and the second point cloud feature, the point cloud segmentation result of the target point cloud data is determined, so that the first point Yun Te determined based on voxel division and the second point cloud feature obtained by point cloud feature extraction can be combined, the point cloud segmentation result is determined, the requirement on the voxel size is not high, the voxel size of the point cloud is larger at the position where the point cloud is more sparse, under the condition that the point cloud segmentation accuracy is ensured, the calculation amount is greatly reduced, and the calculation time is shortened.
Fig. 3 is a flowchart of a point cloud segmentation method according to another embodiment of the disclosure. As shown in fig. 3, the method comprises the steps of:
And step 306, weighting corresponding dimension characteristics of the point characteristics in the splicing point cloud characteristics according to the weight of each dimension to obtain the processed point cloud characteristics.
In this embodiment of the disclosure, the electronic device may input the stitched point cloud feature into the attention mechanism network, and obtain the processed point cloud feature output by the attention mechanism network.
The method comprises the steps of determining the weight of each dimension of the point feature in the point cloud feature to be spliced, reducing the weight of the non-important feature in the point cloud feature, and increasing the weight of the important feature in the point cloud feature, so that the important feature in the point cloud feature can be considered in an emphatic mode, the interference of the non-important feature is avoided, and the classification result of the point in the target point cloud data is determined.
And 307, classifying the processed point cloud characteristics to obtain a classification result of each point in the target point cloud data.
And 308, determining a point cloud segmentation result according to the classification result of each point in the target point cloud data.
In an embodiment of the present disclosure, the classification result of the point may include: probability that a point belongs to each category. For each point, the category to which the point belongs can be determined according to the probability that the point belongs to each category, and then the point cloud segmentation result is determined. For example, the category to which the point belongs is determined as the category to which the point corresponds to the highest probability among the probabilities of the point belonging to the categories.
The target point cloud data is segmented according to the category to which each point in the target point cloud data belongs, and point cloud data of each object in the target point cloud data, namely a point cloud segmentation result, can be obtained. According to the point cloud segmentation result, each object can be modeled respectively, and three-dimensional information, positions and the like of the object can be determined.
In the embodiment of the disclosure, the electronic device may input the processed point cloud features into the fully-connected network, and obtain a classification result of each point output by the fully-connected network.
It should be noted that, the detailed descriptions of step 301, step 302 and step 303 may refer to the corresponding descriptions of step 101, step 102 and step 103 in the embodiment shown in fig. 1, and a repeated description is not provided here.
According to the point cloud segmentation method, target point cloud data are obtained, wherein the target point cloud data are obtained by scanning target objects around a vehicle through laser beams; performing cylinder voxel division and feature extraction on the target point cloud data to obtain first point cloud features of the target point cloud data; performing point cloud feature extraction on the target point cloud data to acquire a second point cloud feature of the target point cloud data; splicing the first point Yun Te and the second point cloud feature to obtain a spliced point cloud feature; determining the weight of each dimension of the point feature in the spliced point cloud feature; according to the weight of each dimension, carrying out weighting processing on corresponding dimension features of the point feature in the splicing point cloud feature to obtain processed point cloud features; classifying the processed point cloud characteristics to obtain a classification result of each point in the target point cloud data; according to the classification result of each point in the target point cloud data, the point cloud segmentation result is determined, so that the first point Yun Te determined based on voxel division and the second point cloud feature obtained by point cloud feature extraction can be combined to determine the point cloud segmentation result, the requirement on the voxel size is low, the voxel size of the more sparse point cloud is larger, the calculation amount is greatly reduced, and the calculation time is shortened under the condition that the point cloud segmentation accuracy is ensured.
Fig. 4 is a schematic structural diagram of a point cloud segmentation apparatus according to an embodiment of the disclosure.
As shown in fig. 4, the point cloud segmentation apparatus may include: an acquisition module 401, a first processing module 402, a second processing module 403 and a determination module 404.
The acquiring module 401 is configured to acquire target point cloud data, where the target point cloud data is data obtained by scanning a target object around a vehicle with a laser beam;
a first processing module 402, configured to perform cylinder voxel division and feature extraction on the target point cloud data, and obtain a first point cloud feature of the target point cloud data;
a second processing module 403, configured to perform point cloud feature extraction on the target point cloud data, and obtain a second point cloud feature of the target point cloud data;
a determining module 404, configured to determine a point cloud segmentation result of the target point cloud data according to the first point Yun Te and the second point cloud feature.
In one embodiment of the present disclosure, the first processing module 402 includes: the device comprises a conversion unit, a dividing unit, a first determining unit and a second determining unit; the conversion unit is used for carrying out coordinate system conversion on the target point cloud data to obtain first point cloud data under a cylindrical coordinate system; the dividing unit is used for carrying out voxel division on the first point cloud data to obtain a plurality of voxels; the voxel is a fan-ring cylinder voxel, and the upper bottom surface and the lower bottom surface of the fan-ring cylinder voxel are in a fan-ring shape; the first determining unit is used for determining voxel characteristics of a plurality of voxels according to local point cloud data in the plurality of voxels; the second determining unit is configured to determine the first point cloud feature according to voxel features of the voxels and the first point cloud data.
In an embodiment of the present disclosure, the first determining unit is specifically configured to, for each voxel, determine voxel data of the voxel according to local point cloud data located within the voxel in the first point cloud data; and carrying out voxel characteristic extraction processing on the voxel data of the voxels to obtain the voxel characteristics of the voxels.
In an embodiment of the present disclosure, the second determining unit is specifically configured to determine center point coordinate information of a plurality of voxels; for each point in the first point cloud data, determining a point feature of the point according to center point coordinate information of a plurality of peripheral voxels of the point, voxel features of the plurality of peripheral voxels, and coordinate information of the point in the first point cloud data; and determining the first point cloud characteristics according to the point characteristics of the points and the point data of the points in the target point cloud data.
In an embodiment of the present disclosure, the second determining unit is specifically configured to, for each point in the first point cloud data, determine a weight of a voxel characteristic of a plurality of peripheral voxels according to center point coordinate information of a plurality of peripheral voxels of the point and coordinate information of the point in the first point cloud data; and determining the point characteristics of the points according to the voxel characteristics of the peripheral voxels and the corresponding weights.
In an embodiment of the present disclosure, the determining module 404 is specifically configured to perform a stitching process on the first point Yun Te and the second point cloud feature to obtain a stitched point cloud feature; determining the weight of each dimensionality of the point feature in the spliced point cloud feature; according to the weight of each dimension, carrying out weighting processing on the corresponding dimension feature of the point feature in the spliced point cloud feature to obtain a processed point cloud feature; classifying the processed point cloud characteristics to obtain a classification result of each point in the target point cloud data; and determining the point cloud segmentation result according to the classification result of each point in the target point cloud data.
The point cloud segmentation device comprises a laser beam scanning device, a point cloud segmentation device and a point cloud segmentation control device, wherein the laser beam scanning device is used for scanning a target object around a vehicle; performing cylinder voxel division and feature extraction on the target point cloud data to acquire first point cloud features of the target point cloud data; performing point cloud feature extraction on the target point cloud data to acquire a second point cloud feature of the target point cloud data; according to the first point Yun Te and the second point cloud feature, the point cloud segmentation result of the target point cloud data is determined, so that the first point Yun Te determined based on voxel division and the second point cloud feature obtained by point cloud feature extraction can be combined, the point cloud segmentation result is determined, the requirement on the voxel size is not high, the voxel size of the point cloud is larger at the position where the point cloud is more sparse, under the condition that the point cloud segmentation accuracy is ensured, the calculation amount is greatly reduced, and the calculation time is shortened.
According to a third aspect of an embodiment of the present disclosure, there is also provided a vehicle including: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to: the steps of the point cloud segmentation method as described above are implemented.
According to a fourth aspect of the embodiments of the present disclosure, there is also provided an electronic apparatus, including: a processor; a memory for storing processor executable instructions, wherein the processor is configured to execute the instructions to implement the point cloud segmentation method as described above.
In order to implement the above embodiments, the present disclosure also provides a storage medium.
Wherein the instructions in the storage medium, when executed by a processor of the electronic device, enable the electronic device to perform the method as above.
To implement the above embodiments, the present disclosure also provides a computer program product.
Wherein the computer program product, when executed by a processor of an electronic device, enables the electronic device to perform the method as above.
Fig. 5 is a block diagram illustrating a structure of an electronic device according to an example embodiment. The electronic device shown in fig. 5 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present disclosure.
As shown in fig. 5, the electronic device 1000 includes a processor 111 that can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM) 112 or a program loaded from a Memory 116 into a Random Access Memory (RAM) 113. In the RAM 113, various programs and data necessary for the operation of the electronic apparatus 1000 are also stored. The processor 111, the ROM 112, and the RAM 113 are connected to each other through a bus 114. An Input/Output (I/O) interface 115 is also connected to the bus 114.
The following components are connected to the I/O interface 115: a memory 116 including a hard disk and the like; and a communication section 117 including a Network interface card such as a Local Area Network (LAN) card, a modem, or the like, the communication section 117 performing communication processing via a Network such as the internet; a drive 118 is also connected to the I/O interface 115 as necessary.
In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program embodied on a computer readable medium, the computer program containing program code for performing the method illustrated by the flow chart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication section 117. Which when executed by the processor 111 performs the above-described functions defined in the method of the present disclosure.
In an exemplary embodiment, a storage medium comprising instructions, such as a memory comprising instructions, executable by the processor 111 of the electronic device 1000 to perform the above-described method is also provided. Alternatively, the storage medium may be a non-transitory computer readable storage medium, for example, the non-transitory computer readable storage medium may be a ROM, a Random Access Memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.
In the present disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In contrast, in the present disclosure, a computer-readable signal medium may include a propagated data signal with computer-readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may 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. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
Fig. 6 is a functional block diagram schematic diagram of a vehicle 600 according to an exemplary embodiment of the present disclosure. For example, the vehicle 600 may be a hybrid vehicle, a non-hybrid vehicle, an electric vehicle, a fuel cell vehicle, or other type of vehicle. Vehicle 600 may be an autonomous vehicle, a semi-autonomous vehicle, or a non-autonomous vehicle.
Referring to fig. 6, a vehicle 600 may include various subsystems such as an infotainment system 610, a perception system 620, a decision control system 630, a drive system 640, and a computing platform 650. The vehicle 600 may also include more or fewer subsystems, and each subsystem may include multiple components, among others. In addition, the interconnection between each subsystem and each component of the vehicle 600 may be achieved through wired or wireless means.
In some embodiments, infotainment system 610 may include a communication system, an entertainment system, and a navigation system, among others.
The sensing system 620 may include several sensors for sensing information about the environment surrounding the vehicle 600. For example, the sensing system 620 may include a global positioning system (the global positioning system may be a GPS system, a beidou system, or other positioning system), an Inertial Measurement Unit (IMU), a laser radar, a millimeter-wave radar, an ultrasonic radar, and a camera.
The drive system 640 may include components that provide powered motion to the vehicle 600. In one embodiment, the drive system 640 may include an engine, an energy source, a transmission system, and wheels. The engine may be one or a combination of internal combustion engine, electric motor, air compression engine. The engine is capable of converting energy provided by the energy source into mechanical energy.
Some or all of the functions of the vehicle 600 are controlled by the computing platform 650. Computing platform 650 may include at least one processor 651 and memory 652, processor 651 may execute instructions 653 stored in memory 652.
The processor 651 may be any conventional processor, such as a commercially available CPU. The processor may also include, for example, an image processor (GPU), a Field Programmable Gate Array (FPGA), a System On Chip (SOC), an Application Specific Integrated Circuit (ASIC), or a combination thereof.
The memory 652 may be implemented by any type or combination of volatile or non-volatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.
In addition to instructions 653, memory 652 may store data such as road maps, route information, location, direction, speed, etc. of the vehicle. Memory 652 stores data that may be used by computing platform 650.
In the disclosed embodiment, processor 651 may execute instructions 653 to perform all or some of the steps of the point cloud segmentation method described above.
Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This application is intended to cover any variations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.
It will be understood that the present disclosure is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.
Claims (9)
1. A method of point cloud segmentation, the method comprising:
acquiring target point cloud data, wherein the target point cloud data is obtained by scanning target objects around a vehicle by laser beams;
performing cylinder voxel division and feature extraction on the target point cloud data to acquire a first point cloud feature of the target point cloud data;
performing point cloud feature extraction on the target point cloud data to acquire a second point cloud feature of the target point cloud data;
determining a point cloud segmentation result of the target point cloud data according to the first point Yun Te and the second point cloud feature;
the cylinder voxel division and feature extraction are carried out on the target point cloud data, and the first point cloud feature of the target point cloud data is obtained, and the method comprises the following steps:
converting a coordinate system of the target point cloud data to obtain first point cloud data under a cylindrical coordinate system;
carrying out voxel division on the first point cloud data to obtain a plurality of voxels; the voxel is a fan-ring cylinder voxel, and the upper bottom surface and the lower bottom surface of the fan-ring cylinder voxel are in a fan-ring shape;
determining voxel characteristics of a plurality of voxels according to local point cloud data in the plurality of voxels;
determining the first point cloud feature according to the voxel features of the voxels and the first point cloud data;
determining the first point cloud feature according to the voxel features of the plurality of voxels and the first point cloud data, including:
determining center point coordinate information of a plurality of voxels;
for each point in the first point cloud data, determining a point characteristic of each point according to central point coordinate information of a plurality of peripheral voxels of each point, voxel characteristics of the plurality of peripheral voxels, and the coordinate information of each point in the first point cloud data;
determining the first point cloud characteristics according to the point characteristics of the plurality of points and the point data of the plurality of points in the target point cloud data;
the determining, for each point in the first point cloud data, a point feature of each point according to center point coordinate information of a plurality of peripheral voxels of the each point, voxel features of the plurality of peripheral voxels, and coordinate information of the each point in the first point cloud data includes:
for each point in the first point cloud data, determining the weight of voxel characteristics of a plurality of peripheral voxels according to center point coordinate information of a plurality of peripheral voxels of the each point and the coordinate information of the each point in the first point cloud data;
and determining the point characteristics of each point according to the voxel characteristics of a plurality of peripheral voxels and corresponding weights, wherein the weighted summation is carried out on the voxel characteristics of the plurality of peripheral voxels according to the corresponding weights to obtain the point characteristics of each point.
2. The method of claim 1, wherein determining voxel characteristics of a plurality of the voxels from local point cloud data within the plurality of voxels comprises:
for each voxel, determining voxel data of the voxel according to local point cloud data, located in the voxel, in the first point cloud data;
and carrying out voxel characteristic extraction processing on the voxel data of the voxels to obtain the voxel characteristics of the voxels.
3. The method of claim 1, wherein determining the point cloud segmentation result for the target point cloud data from the first point Yun Te and the second point cloud feature comprises:
splicing the first point Yun Te and the second point cloud feature to obtain a spliced point cloud feature;
determining the weight of each dimension of the point feature in the spliced point cloud feature;
according to the weight of each dimension, carrying out weighting processing on corresponding dimension features of the point feature in the spliced point cloud feature to obtain processed point cloud features;
classifying the processed point cloud characteristics to obtain a classification result of each point in the target point cloud data;
and determining the point cloud segmentation result according to the classification result of each point in the target point cloud data.
4. A point cloud segmentation apparatus, characterized in that the apparatus comprises:
the system comprises an acquisition module, a processing module and a display module, wherein the acquisition module is used for acquiring target point cloud data, and the target point cloud data is obtained by scanning a target object around a vehicle by a laser beam;
the first processing module is used for performing cylinder voxel division and feature extraction on the target point cloud data to acquire a first point cloud feature of the target point cloud data;
the second processing module is used for extracting point cloud characteristics of the target point cloud data to obtain second point cloud characteristics of the target point cloud data;
the determining module is used for determining a point cloud segmentation result of the target point cloud data according to the first point Yun Te and the second point cloud characteristics;
the first processing module comprises: the device comprises a conversion unit, a dividing unit, a first determining unit and a second determining unit;
the conversion unit is used for carrying out coordinate system conversion on the target point cloud data to obtain first point cloud data under a cylindrical coordinate system;
the dividing unit is used for carrying out voxel division on the first point cloud data to obtain a plurality of voxels; the voxel is a fan-ring cylinder voxel, and the upper bottom surface and the lower bottom surface of the fan-ring cylinder voxel are in a fan-ring shape;
the first determining unit is used for determining voxel characteristics of a plurality of voxels according to local point cloud data in the plurality of voxels;
the second determining unit is configured to determine the first point cloud feature according to voxel features of a plurality of voxels and the first point cloud data;
the second determination unit is specifically configured to,
determining center point coordinate information of a plurality of voxels;
for each point in the first point cloud data, determining a point characteristic of each point according to central point coordinate information of a plurality of peripheral voxels of each point, voxel characteristics of the plurality of peripheral voxels, and the coordinate information of each point in the first point cloud data;
determining the first point cloud characteristics according to the point characteristics of the plurality of points and the point data of the plurality of points in the target point cloud data;
the determining, for each point in the first point cloud data, a point feature of each point according to center point coordinate information of a plurality of peripheral voxels of the each point, voxel features of the plurality of peripheral voxels, and coordinate information of the each point in the first point cloud data includes:
for each point in the first point cloud data, determining the weight of voxel characteristics of a plurality of peripheral voxels according to center point coordinate information of a plurality of peripheral voxels of the each point and the coordinate information of the each point in the first point cloud data;
and determining the point characteristics of each point according to the voxel characteristics of a plurality of peripheral voxels and corresponding weights, wherein the weighted summation is carried out on the voxel characteristics of the plurality of peripheral voxels according to the corresponding weights to obtain the point characteristics of each point.
5. The apparatus according to claim 4, characterized in that the first determination unit is specifically configured to,
for each voxel, determining voxel data of the voxel according to local point cloud data located in the voxel in the first point cloud data;
and carrying out voxel characteristic extraction processing on the voxel data of the voxels to obtain the voxel characteristics of the voxels.
6. The apparatus of claim 4, wherein the means for determining is specifically configured to,
splicing the first point Yun Te and the second point cloud feature to obtain a spliced point cloud feature;
determining the weight of each dimensionality of the point feature in the spliced point cloud feature;
according to the weight of each dimension, carrying out weighting processing on the corresponding dimension feature of the point feature in the spliced point cloud feature to obtain a processed point cloud feature;
classifying the processed point cloud characteristics to obtain a classification result of each point in the target point cloud data;
and determining the point cloud segmentation result according to the classification result of each point in the target point cloud data.
7. A vehicle, characterized by comprising:
a processor;
a memory for storing processor-executable instructions;
wherein the processor is configured to:
a step of implementing the point cloud segmentation method of any one of claims 1 to 3.
8. An electronic device, comprising:
a processor;
a memory for storing the processor-executable instructions;
wherein the processor is configured to execute the instructions to implement the point cloud segmentation method of any of claims 1 to 3.
9. A non-transitory computer readable storage medium, instructions in which when executed by a processor, enable the processor to perform the point cloud segmentation method of any of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211357371.6A CN115471513B (en) | 2022-11-01 | 2022-11-01 | Point cloud segmentation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211357371.6A CN115471513B (en) | 2022-11-01 | 2022-11-01 | Point cloud segmentation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115471513A CN115471513A (en) | 2022-12-13 |
| CN115471513B true CN115471513B (en) | 2023-03-31 |
Family
ID=84337223
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211357371.6A Active CN115471513B (en) | 2022-11-01 | 2022-11-01 | Point cloud segmentation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115471513B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115082885A (en) * | 2022-06-27 | 2022-09-20 | 深圳见得空间科技有限公司 | Point cloud target detection method, device, equipment and storage medium |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110807774B (en) * | 2019-09-30 | 2022-07-12 | 九天创新(广东)智能科技有限公司 | Point cloud classification and semantic segmentation method |
| CN111489358B (en) * | 2020-03-18 | 2022-06-14 | 华中科技大学 | Three-dimensional point cloud semantic segmentation method based on deep learning |
| CN114764141A (en) * | 2020-12-30 | 2022-07-19 | 中寰卫星导航通信有限公司 | Laser radar point clustering method and device and storage medium |
| CN113378854A (en) * | 2021-06-11 | 2021-09-10 | 武汉大学 | Point cloud target detection method integrating original point cloud and voxel division |
| CN114488185A (en) * | 2022-03-01 | 2022-05-13 | 山东大学 | Robot navigation system method and system based on multi-line laser radar |
-
2022
- 2022-11-01 CN CN202211357371.6A patent/CN115471513B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115082885A (en) * | 2022-06-27 | 2022-09-20 | 深圳见得空间科技有限公司 | Point cloud target detection method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115471513A (en) | 2022-12-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11594011B2 (en) | Deep learning-based feature extraction for LiDAR localization of autonomous driving vehicles | |
| WO2020154973A1 (en) | Lidar localization using rnn and lstm for temporal smoothness in autonomous driving vehicles | |
| US11875521B2 (en) | Self-occlusion masks to improve self-supervised monocular depth estimation in multi-camera settings | |
| WO2024012211A1 (en) | Autonomous-driving environmental perception method, medium and vehicle | |
| CN116027951B (en) | Visualization method and device for vehicle acquisition data and storage medium | |
| CN115223015B (en) | Model training method, image processing method, device and vehicle | |
| CN116486377B (en) | Method and device for generating drivable area | |
| CN115471513B (en) | Point cloud segmentation method and device | |
| CN115718304A (en) | Target object detection method, target object detection device, vehicle and storage medium | |
| CN116343174A (en) | Target detection method, device, vehicle and storage medium | |
| CN116168362A (en) | Pre-training method and device for vehicle perception model, electronic equipment and vehicle | |
| WO2021212297A1 (en) | Systems and methods for distance measurement | |
| CN114549610A (en) | Point cloud data processing method and related device | |
| CN116385825B (en) | Model joint training method and device and vehicle | |
| CN116678423B (en) | Multisource fusion positioning method, multisource fusion positioning device and vehicle | |
| CN114842458B (en) | Obstacle detection method, obstacle detection device, vehicle, and storage medium | |
| CN116259029B (en) | Target detection method and device and vehicle | |
| CN116659529B (en) | Data detection method, device, vehicle and storage medium | |
| CN116863429B (en) | Training method of detection model, and determination method and device of exercisable area | |
| CN114821543B (en) | Obstacle detection method, obstacle detection device, vehicle, and storage medium | |
| CN116740681B (en) | Target detection method, device, vehicle and storage medium | |
| CN117128976B (en) | Method and device for acquiring road center line, vehicle and storage medium | |
| CN116503482B (en) | Vehicle position acquisition method and device and electronic equipment | |
| US11620831B2 (en) | Register sets of low-level features without data association | |
| US20230296402A1 (en) | Method, System, and Computer Program Product for Iterative Warping of Maps for Autonomous Vehicles and Simulators |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |