CN113366532B - Point cloud based segmentation processing method and device, computer equipment and storage medium - Google Patents
Point cloud based segmentation processing method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN113366532B CN113366532B CN201980037714.0A CN201980037714A CN113366532B CN 113366532 B CN113366532 B CN 113366532B CN 201980037714 A CN201980037714 A CN 201980037714A CN 113366532 B CN113366532 B CN 113366532B
- Authority
- CN
- China
- Prior art keywords
- point
- point cloud
- points
- sub
- data
- 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
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
Abstract
A point cloud-based segmentation processing method comprises the following steps: acquiring point cloud data, wherein the point cloud data comprises point data corresponding to a plurality of points; dividing the point cloud data into a plurality of sub-point clouds according to the point data; generating a segmentation threshold corresponding to the sub-point cloud based on the point data corresponding to the plurality of points in the sub-point cloud; traversing a plurality of points in the sub-point cloud, and determining points of which vertical axis coordinates are smaller than the segmentation threshold in the point data as ground points; and counting the ground points in the plurality of sub-point clouds to obtain a ground point set corresponding to the point cloud data.
Description
Technical Field
The application relates to a point cloud based segmentation processing method and device, computer equipment, storage medium and vehicle.
Background
With the development of artificial intelligence and machine vision in the field of intelligent motion control, particularly in the field of unmanned driving, laser sensors are used more and more. The laser sensor has the advantages of high speed, high precision, strong photoelectric interference resistance and the like, can accurately measure the position of an object in a three-dimensional space, and can control various intelligent moving machines such as unmanned vehicles and the like according to data acquired by the laser sensor. Data acquired by the laser sensor are point cloud data, the point cloud data are generally divided on the ground in a traditional mode, and vehicle control is performed by using the divided point data.
However, the inventor realizes that the conventional ground segmentation method based on the point cloud data is slow in speed, and is not beneficial to quickly controlling according to the collected point cloud data in the unmanned driving process. Therefore, how to improve the segmentation efficiency of the point cloud data becomes a technical problem to be solved at present.
Disclosure of Invention
According to various embodiments disclosed in the application, a point cloud-based segmentation processing method, a point cloud-based segmentation processing device, a computer device, a storage medium and a vehicle are provided.
A point cloud-based segmentation processing method comprises the following steps:
acquiring point cloud data, wherein the point cloud data comprises point data corresponding to a plurality of points;
dividing the point cloud data into a plurality of sub-point clouds according to the point data;
generating a segmentation threshold corresponding to the sub-point cloud based on the point data corresponding to the plurality of points in the sub-point cloud;
traversing a plurality of points in the sub-point cloud, and determining points of which vertical axis coordinates are smaller than the segmentation threshold in the point data as ground points; and
and counting the ground points in the sub-point clouds to obtain a ground point set corresponding to the point cloud data.
A point cloud based segmentation processing apparatus comprising:
the point cloud acquisition module is used for acquiring point cloud data, and the point cloud data comprises point data corresponding to a plurality of points;
the point cloud dividing module is used for dividing the point cloud data into a plurality of sub-point clouds according to the point data;
the threshold generating module is used for generating a segmentation threshold corresponding to the sub-point cloud based on the point data corresponding to the plurality of point points in the sub-point cloud; traversing a plurality of points in the sub-point cloud, and determining points of which vertical axis coordinates are smaller than the segmentation threshold in the point data as ground points; and
and the ground point counting module is used for counting ground points in the sub-point clouds to obtain a ground point set corresponding to the point cloud data.
A computer device comprising a memory and one or more processors, the memory having stored therein computer-readable instructions that, when executed by the processors, cause the one or more processors to perform the steps of:
acquiring point cloud data, wherein the point cloud data comprises point data corresponding to a plurality of points;
dividing the point cloud data into a plurality of sub-point clouds according to the point data;
generating a segmentation threshold corresponding to the sub-point cloud based on the point data corresponding to the plurality of points in the sub-point cloud;
traversing a plurality of points in the sub-point cloud, and determining points of which vertical axis coordinates are smaller than the segmentation threshold in the point data as ground points; and
and counting the ground points in the sub-point clouds to obtain a ground point set corresponding to the point cloud data.
One or more non-transitory computer-readable storage media storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of:
acquiring point cloud data, wherein the point cloud data comprises point data corresponding to a plurality of points;
dividing the point cloud data into a plurality of sub-point clouds according to the point data;
generating a segmentation threshold corresponding to the sub-point cloud based on the point data corresponding to the plurality of points in the sub-point cloud;
traversing a plurality of points in the sub-point cloud, and determining points of which vertical axis coordinates are smaller than the segmentation threshold in the point data as ground points; and
and counting the ground points in the sub-point clouds to obtain a ground point set corresponding to the point cloud data.
A vehicle comprises the steps of executing the point cloud-based segmentation processing method.
The details of one or more embodiments of the application are set forth in the accompanying drawings and the description below. Other features and advantages of the application will be apparent from the description and drawings, and from the claims.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
Fig. 1 is a diagram of an application scenario of a point cloud-based segmentation processing method according to one or more embodiments.
FIG. 2 is a schematic flow diagram of a point cloud based segmentation processing method in accordance with one or more embodiments.
FIG. 3 is a flow diagram illustrating a step of generating segmentation thresholds corresponding to sub-point clouds based on point data corresponding to multiple points in the sub-point clouds in accordance with one or more embodiments.
FIG. 4 is a block diagram of a point cloud based segmentation processing device in accordance with one or more embodiments.
FIG. 5 is a block diagram of a computer device in accordance with one or more embodiments.
Detailed Description
In order to make the technical solutions and advantages of the present application more clearly understood, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
The point cloud based segmentation processing method provided by the application can be applied to the application environment shown in fig. 1. The laser sensor 102 communicates with a computer device 104 over a network, and the computer device 104 may be an on-board computer device. The laser sensor 102 may collect environmental data around vehicles, including autonomous cars, bicycles, aircrafts, and the like, and the collected environmental data may be specifically point cloud data. The computer device 104 obtains point cloud data, which includes point data corresponding to a plurality of points. The computer device 104 divides the point cloud data into a plurality of sub-point clouds according to the point data, and generates segmentation thresholds corresponding to the sub-point clouds based on the point data corresponding to the plurality of points in the sub-point clouds. The computer device 104 traverses a plurality of points in the sub-point cloud, and determines a point of which the vertical axis coordinate in the point data is smaller than the segmentation threshold value as a ground point; and counting the ground points in the plurality of sub-point clouds to obtain a ground point set corresponding to the point cloud data. The laser sensor 102 may be a laser sensor mounted on a vehicle, and may specifically include a laser radar, a laser scanner, and the like. The computer device 104 may be a vehicle-mounted computer device.
In one embodiment, as shown in fig. 2, a point cloud-based segmentation processing method is provided, which is exemplified by being applied to the computer device 104 in fig. 1, and includes the following steps:
The point cloud data may be data collected by a laser sensor. The laser sensor may be a laser sensor mounted on a vehicle such as a vehicle, and specifically may include a laser radar, a laser scanner, and the like. The laser sensor may be disposed on the unmanned vehicle for collecting environmental data about the vehicle. The laser sensor may emit a detection signal, such as a laser beam, and compare the reflected signal with the detection signal to obtain environmental data around the vehicle. The environmental data may specifically be data in a point cloud format. The point cloud data is a collection of point data corresponding to a plurality of points on the surface of the object, which is recorded in the form of points by scanning the object.
The computer equipment can acquire multi-frame point cloud data, and the point cloud data can be acquired by a laser sensor or pre-stored in the computer equipment. The computer device can respectively process each frame of point cloud data. Each frame of point cloud data may include point data corresponding to a plurality of points. The dot data may specifically include at least one of three-dimensional coordinates, laser reflection intensity, color information, and the like of the dot correspondence. The three-dimensional coordinates may be coordinates of the point in a cartesian coordinate system, and specifically include a horizontal axis coordinate (x axis), a vertical axis coordinate (y axis), and a vertical axis coordinate (z axis) of the point in the cartesian coordinate system. The Cartesian coordinate system is a three-dimensional space coordinate system established by taking the laser sensor as an origin, the three-dimensional space coordinate system comprises a transverse axis, a longitudinal axis and a vertical axis, and the three-dimensional space coordinate system established by taking the laser sensor as the origin meets the right-hand rule.
The computer device may divide the point cloud data into a plurality of sub-point clouds according to the point data corresponding to each point in the point cloud data. Specifically, the computer device may record the position of the laser sensor as an origin, and divide a horizontal plane in which the laser sensor is located into a plurality of grids. The computer equipment can determine the grids to which the points belong according to the horizontal axis coordinates and the vertical axis coordinates corresponding to the points, and a set formed by all the points in the same grid is recorded as a sub-point cloud. The sub-point clouds are a subset of the corresponding point clouds. After the point cloud data are divided into the plurality of sub-point clouds by the computer equipment, the sub-point clouds can be segmented one by a single thread, and the sub-point clouds can be segmented in parallel by multiple threads, so that the point cloud segmentation efficiency is improved. Compared with the traditional ground fitting mode, the method is only suitable for flat ground, the point cloud data are divided into a plurality of sub-point clouds for segmentation, the method can be suitable for uneven ground, and the universality of point cloud data segmentation is effectively improved.
The computer device may use one of a plurality of division methods to divide the horizontal plane on which the laser sensor is located into a plurality of grids. For example, the computer device may divide the horizontal plane in which the laser sensor is located into a plurality of square grids of equal length. The side length of the square grid can be preset according to actual requirements, for example, the horizontal plane can be divided into a plurality of square grids of 1 meter by 1 meter.
In one embodiment, the computer device may also establish a polar coordinate system with the laser sensor as the origin. The computer equipment can map the three-dimensional coordinates corresponding to the points in the point cloud data into the polar coordinate system to obtain the distribution result of the point cloud data in the polar coordinate system. The computer device may generate a polar grid map from the polar coordinate system, dividing the polar coordinate system into a plurality of equal-circumference sectors. The computer device may divide each sector into a plurality of grids at equal distances, resulting in a plurality of polar grids. The computer device may count the points in each polar coordinate grid to obtain a set of corresponding points in the polar coordinate grid, and the computer device may record the set of corresponding points in each polar coordinate grid as a sub-point cloud corresponding to the polar coordinate grid.
In this embodiment, the point cloud data generally has the characteristics of close density and far sparseness, that is, the points closer to the laser sensor are relatively dense, and the points farther from the laser sensor are relatively sparse. And the area of the grid close to the origin in the polar coordinate grid is smaller, and the area of the grid far away from the origin is larger. The computer equipment divides the point cloud data into sub-point clouds corresponding to the polar coordinate grids, the number of the midpoints of each sub-point cloud is balanced, and the accuracy of point cloud segmentation is effectively improved.
And step 206, generating a segmentation threshold corresponding to the sub-point cloud based on the point data corresponding to the plurality of points in the sub-point cloud.
The computer equipment can acquire point data corresponding to each point from the sub-point cloud and generate a segmentation threshold corresponding to the sub-point cloud according to the point data corresponding to all the points. The segmentation threshold refers to a threshold for segmenting points in the sub-point cloud into ground points and non-ground points, and may be a maximum height value generated from the point data. As can be appreciated, the computer device can process each of the plurality of sub-point clouds independently, each sub-point cloud having a respective corresponding segmentation threshold. The segmentation threshold between different sub-point clouds may be the same or different.
Specifically, the computer device may obtain a vertical axis coordinate corresponding to a midpoint of the sub-point cloud, and the vertical axis coordinate may reflect a height of the point in a vertical axis direction. The computer device can evaluate the height of the ground in the grid according to the corresponding vertical axis coordinate to obtain a ground height coordinate, wherein the ground height coordinate can be the vertical axis coordinate corresponding to the ground highest point in the grid corresponding to the sub-point cloud. The computer equipment can record the coordinate value of the ground height coordinate as a segmentation threshold corresponding to the sub-point cloud, and perform segmentation processing on the points in the sub-point cloud by using the segmentation threshold.
And step 208, traversing a plurality of points in the sub-point cloud, and determining points of which the vertical axis coordinates in the point data are smaller than the segmentation threshold value as ground points.
And step 210, counting ground points in the plurality of sub-point clouds to obtain a ground point set corresponding to the point cloud data.
The computer equipment can traverse a plurality of points in the sub-point cloud, sequentially compare the vertical axis coordinate corresponding to each point with the segmentation threshold, judge whether the vertical axis coordinate is smaller than the segmentation threshold, and label the points according to the comparison result between the vertical axis coordinate and the segmentation threshold. Specifically, the computer device compares the vertical axis coordinates of the points one by one with the segmentation threshold. When the vertical axis coordinate of the point is less than the segmentation threshold, the computer device may mark the point as a ground point. When the vertical axis coordinate of the point is greater than or equal to the segmentation threshold, the computer device may mark the point as a non-ground point. And traversing each point in the sub-point cloud by the computer equipment, sequentially labeling the points, and segmenting the sub-point cloud until all the points in the sub-point cloud are segmented and labeled.
The computer device may obtain the ground points included in the sub-point cloud after the traversal of the points in the sub-point cloud is finished. The computer device may perform statistical analysis on the ground points corresponding to each of the plurality of sub-point clouds after performing segmentation processing on each of the plurality of sub-point clouds. The computer device may generate a ground point set according to the plurality of ground points corresponding to the plurality of divided sub-point clouds. The ground point set is a set consisting of ground points in the point cloud data collected by the laser sensor, and points in the point cloud data except the ground point set are non-ground points, so that the segmentation between the ground points and the non-ground points of the point cloud data is completed.
In this embodiment, after the point cloud data is acquired by the computer device, the point cloud data is divided into a plurality of sub-point clouds according to the point data in the point cloud data, and each sub-point cloud is relatively independent, which is beneficial to performing point cloud segmentation on uneven ground. The computer equipment can call multiple threads to process multiple sub-point clouds in parallel, and the efficiency of point cloud segmentation is improved. The computer equipment can generate a segmentation threshold corresponding to the sub-point cloud based on the point data in the sub-point cloud, traverse a plurality of points in the sub-point cloud, determine a point of which the vertical axis coordinate in the point data is smaller than the segmentation threshold as a ground point, and rapidly segment the points in the sub-point cloud. The computer device counts the ground points in the plurality of sub-point clouds to obtain a ground point set corresponding to the point cloud data, so that the point cloud data is segmented, and the segmentation efficiency of the point cloud data is effectively improved.
In one embodiment, as shown in fig. 3, generating a segmentation threshold corresponding to a sub-point cloud based on point data corresponding to a plurality of points in the sub-point cloud includes:
And step 306, generating a segmentation threshold corresponding to the sub-point cloud based on the ground height coordinate and a preset height threshold.
The laser sensor can establish a three-dimensional coordinate system by taking the position as an origin, point data acquired by the laser sensor comprises corresponding three-dimensional coordinates, and the three-dimensional coordinates are position coordinates pointed in the three-dimensional coordinate system established by taking the laser sensor as the origin. The computer device can obtain three-dimensional coordinates corresponding to a plurality of points from point data included in the sub-point cloud, and extract vertical axis coordinates corresponding to the points from the three-dimensional coordinates. The computer equipment can evaluate the ground height according to the obtained vertical axis coordinates to obtain the ground height coordinates. The ground height coordinate includes only the coordinate value on the vertical axis in the three-dimensional coordinate system with the laser sensor as the origin.
The computer device can also obtain a preset height threshold corresponding to the laser sensor, and the height threshold can be used for adjusting the ground height coordinate. The height threshold value can be preset by a user according to actual requirements, specifically can be set according to the installation position, angle and the like of the laser sensor, and can be preset according to the laser sensor. The height threshold may be a constant, in particular a constant between 10 cm and 20 cm, for example 15 cm.
The computer device can generate a segmentation threshold corresponding to the sub-point cloud according to the estimated ground height coordinate and the height threshold corresponding to the laser sensor. Specifically, the computer device may calculate a sum of the ground height coordinate and the height threshold, and use the sum of the ground height coordinate and the height threshold as the segmentation threshold corresponding to the sub-point cloud. The segmentation threshold is the maximum value of the ground height in the corresponding grid of the sub-point cloud on the vertical axis, and the computer equipment can perform segmentation processing on the points in the sub-point cloud according to the segmentation threshold.
In this embodiment, the computer device evaluates the ground height coordinate according to the vertical axis coordinates corresponding to the multiple points in the sub-point cloud, and obtains a preset height threshold corresponding to the laser sensor to adjust the ground height coordinate. The computer equipment can generate a segmentation threshold value corresponding to the sub point cloud according to the ground height coordinate and the height threshold value, and the generated segmentation threshold value can reflect the maximum height of the ground in the grid corresponding to the sub point cloud more accurately. And the segmentation processing of the corresponding sub-point cloud is carried out according to the generated segmentation threshold, so that the accuracy of point cloud segmentation is effectively improved.
In one embodiment, estimating a ground height coordinate from a plurality of vertical axis coordinates comprises: sequencing the plurality of points according to the vertical axis coordinate to obtain a point sequence; acquiring points at preset positions in the point sequence and recording the points as points to be evaluated; and calculating the average value of the vertical axis coordinates corresponding to the points to be evaluated based on the vertical axis coordinates corresponding to the points to be evaluated as the ground height coordinates corresponding to the sub-point clouds.
And the computer equipment can sort the points in the sub-point cloud according to the size of the corresponding vertical axis coordinate value to obtain the point sequence corresponding to the sub-point cloud. The computer equipment can sort the points from small to large according to the vertical axis coordinate to obtain an ascending point sequence. The computer equipment can also sequence the points from big to small to obtain the descending point sequence.
In one embodiment, the computer device may traverse the points in the sub-point cloud according to the sequence of points until the alignment result is different from the last alignment result. For example, when the point sequence is an ascending point sequence, the computer device compares the vertical axis coordinate with the segmentation threshold in order from small to large. When the vertical axis coordinate is greater than or equal to the segmentation threshold value and the vertical axis coordinate is smaller than the segmentation threshold value in the last comparison result, the computer equipment can finish traversing and directly mark the subsequent points as non-ground points. The computer equipment traverses the points according to the point sequence without completely traversing all the points, thereby effectively saving the resources of the computer equipment and improving the segmentation efficiency of the point cloud data.
The computer device may obtain a point at a predetermined position from the sequence of points and record the point as a point to be evaluated. The preset position is preset by a user according to actual requirements, and the preset position can be obtained according to a large amount of point cloud data and ground point statistics in the point cloud data. The preset position may specifically include a position and a number in a point sequence preset by a user.
For example, the preset position may be a number preset in ascending order from the smallest end of the vertical axis coordinate in the point sequence. The preset number may be a constant, for example the preset number may be set to 50. The preset number may also be a variable, for example, the preset number may be a proportional number in the sub-point cloud, for example, the preset number may be 5% to 10% of the total number of the midpoints in the sub-point cloud. The computer equipment can obtain points accounting for 5% of the total number of the points in ascending order from the minimum end of the point sequence and record the points as the points to be evaluated.
The computer device can obtain the vertical axis coordinate corresponding to the point to be evaluated, and calculate the average value of the vertical axis coordinates corresponding to the multiple points to be evaluated. The computer equipment can record the average value of the vertical axis coordinates corresponding to the multiple points to be evaluated as the ground height coordinates corresponding to the sub-point clouds.
In this embodiment, the computer device orders the points in the sub-point cloud according to the vertical axis coordinates, only part of the points are required to be acquired from the point sequence and marked as the points to be evaluated, and the ground height coordinates are calculated according to the vertical axis coordinates corresponding to the points to be evaluated, so that the calculation amount of the computer device is reduced, the calculation resources of the computer device are saved, and the efficiency of calculating the ground height coordinates is improved. The average value of the vertical axis coordinates corresponding to the point to be evaluated at the preset position can accurately reflect the height of the sub-point cloud corresponding to the ground in the grid, and the accuracy of the segmentation threshold is improved.
In one embodiment, the point cloud data is collected from a laser sensor, the laser sensor includes a first sensor and a second sensor, and the method further includes: acquiring a position relation between a second sensor and a first sensor; and mapping the point cloud data corresponding to the second sensor to a coordinate system corresponding to the first sensor according to the position relation.
In order to collect point cloud data around the vehicle from multiple directions, a plurality of laser sensors can be installed on the vehicle. For example, a plurality of laser sensors may be provided at the roof, the head, the tail, the left side, the right side, and the like of the vehicle. In the plurality of laser sensors, a first sensor and a second sensor may be included. Wherein, the first sensor is the main sensor. For example, a laser sensor mounted on the roof of a vehicle may be referred to as a first sensor. The first sensor includes and includes only one laser sensor. The second sensor is referred to as a secondary sensor. For example, when the laser sensor mounted on the roof of the vehicle is the first sensor, at least one of the laser sensors mounted on the head, tail, left side, and right side of the vehicle body may be referred to as the second sensor. The second sensor may comprise one or more laser sensors. The point cloud data collected by different laser sensors are usually collected by using the corresponding laser sensor as an origin.
After the first sensor and the second sensor are installed on the vehicle, the positions of the first sensor and the second sensor relative to the vehicle are fixed, and the position between the first sensor and the second sensor is also fixed. The computer device can acquire the position relation between the second sensor and the first sensor according to the sensor identifications of the first sensor and the second sensor.
The computer equipment can map the point cloud data acquired by the second sensor to the coordinate system corresponding to the first sensor according to the position relation. Specifically, the computer device may establish a mapping relationship between a first coordinate system corresponding to the first sensor and a second coordinate system corresponding to the second sensor according to a position relationship between the first sensor and the second sensor. The computer equipment can map the point cloud data acquired by the second sensor to the first coordinate system corresponding to the first sensor according to the mapping relation between the first coordinate system and the second coordinate system, so that coordinate system conversion is realized.
In this embodiment, the computer device maps the point cloud data acquired by the second sensor to the coordinate system corresponding to the first sensor by acquiring the position relationship between the first sensor and the second sensor, thereby implementing the fusion of the point cloud data acquired by the plurality of sensors. Compared with the mode that the point cloud data corresponding to the sensors cannot be fused in the traditional image segmentation, the point cloud data collected by the sensors can be fused and then segmented in the embodiment through the position relation between the first sensor and the second sensor, the point cloud data collected by the sensors can be integrated, the comprehensive vehicle multi-directional point cloud data is facilitated, and the accuracy of vehicle control according to the point cloud data is improved.
In one embodiment, before generating the segmentation threshold corresponding to the sub-point cloud based on the point data corresponding to the plurality of points in the sub-point cloud, the method further includes: acquiring a sensor identifier corresponding to the point cloud data; acquiring a corresponding filtering threshold according to the sensor identifier; and screening the point data based on the filtering threshold value to obtain the screened point data.
The computer device may also preprocess the point data included in the sub-point cloud before generating the segmentation threshold corresponding to the sub-point cloud according to the point data corresponding to the point. The preprocessing may specifically include data cleaning of noisy data in the point data. Specifically, the computer device may obtain the sensor identifier corresponding to the laser sensor that collects the point cloud data when obtaining the point cloud data collected by the laser sensor. The computer device can obtain the filtering threshold corresponding to the sensor identifier according to the sensor identifier.
The filtering threshold refers to at least one of a maximum value or a minimum value used for filtering noise data in the point data. The filtering threshold is a coordinate value on the vertical axis of the three-dimensional space coordinate system. The filtering threshold corresponding to the laser sensor can be preset by a user according to actual requirements, and the filtering threshold can be determined according to the installation position of the laser sensor on the vehicle and the height of the laser sensor from the ground. Different laser sensors may correspond to the same or different filtering thresholds. The computer device may pre-associate and store the sensor identifier and the corresponding filtering threshold in a database corresponding to the computer device, and the computer device may obtain the filtering threshold corresponding to the sensor identifier according to an association relationship between the sensor identifier and the filtering threshold.
The computer device can screen the point data according to the filtering threshold value, delete the noise data which do not meet the filtering threshold value from the point data, and generate the segmentation threshold value corresponding to the sub-point cloud by using the screened point data. For example, when the filter threshold includes a maximum filter value and a minimum filter value, the computer device may compare the vertical axis coordinate of the midpoint of the sub-point cloud to the maximum filter value and the minimum filter value. And when the vertical axis coordinate is larger than the minimum filtering value and smaller than the maximum filtering value, determining that the corresponding point data meets the filtering threshold value. When the vertical axis coordinate is less than or equal to the minimum filtering value, or greater than or equal to the maximum filtering value, it is determined that the corresponding point data does not satisfy the filtering threshold value, and belongs to the noise data, and the computer device may delete the corresponding point data.
In this embodiment, before generating the segmentation threshold according to the point data, the computer device may obtain a filtering threshold corresponding to the laser sensor, filter the point data in the sub-point cloud according to the filtering threshold, and generate the segmentation threshold corresponding to the sub-point cloud by using the filtered point data, so as to avoid the influence of noise data in the sub-point cloud on the generation of the segmentation threshold, effectively improve the accuracy of generating the segmentation threshold, and further improve the accuracy of performing segmentation processing on the point cloud data.
It should be understood that although the various steps in the flow charts of fig. 2-3 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 2-3 may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performance of the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternating with other steps or at least some of the sub-steps or stages of other steps.
In one embodiment, as shown in fig. 4, there is provided a point cloud-based segmentation processing apparatus, including: a point cloud obtaining module 402, a point cloud partitioning module 404, a threshold generating module 406, and a ground point statistics module 408, wherein:
a point cloud obtaining module 402, configured to obtain point cloud data, where the point cloud data includes point data corresponding to multiple points.
A point cloud partitioning module 404, configured to partition the point cloud data into a plurality of sub-point clouds according to the point data.
A threshold generating module 406, configured to generate a segmentation threshold corresponding to the sub-point cloud based on point data corresponding to multiple points in the sub-point cloud; and traversing a plurality of points in the sub-point cloud, and determining points of which the vertical axis coordinates in the point data are smaller than the segmentation threshold value as ground points.
The ground point statistics module 408 is configured to count ground points in the plurality of sub-point clouds to obtain a ground point set corresponding to the point cloud data.
In one embodiment, the threshold generating module 406 is further configured to obtain vertical axis coordinates corresponding to a plurality of points from the point data; evaluating a ground height coordinate according to the vertical axis coordinates; and generating a segmentation threshold corresponding to the sub-point cloud based on the ground height coordinate and a preset height threshold.
In one embodiment, the threshold generating module 406 is further configured to sort the plurality of points according to the vertical axis coordinate to obtain a point sequence; acquiring points at preset positions in the point sequence, and recording the points as points to be evaluated; and calculating the average value of the vertical axis coordinates corresponding to the points to be evaluated based on the vertical axis coordinates corresponding to the points to be evaluated as the ground height coordinates corresponding to the sub-point clouds.
In one embodiment, the point cloud data is collected from a laser sensor, the laser sensor comprises a first sensor and a second sensor, and the device further comprises a coordinate system conversion module used for acquiring the position relation between the second sensor and the first sensor; and mapping the point cloud data corresponding to the second sensor to a coordinate system corresponding to the first sensor according to the position relation.
In one embodiment, the point cloud partitioning module 404 is further configured to map the corresponding three-dimensional coordinates to a polar coordinate system; generating a plurality of polar coordinate grids according to the polar coordinate system; and counting the points in the polar coordinate grid to obtain the sub-point cloud corresponding to the polar coordinate grid.
In one embodiment, the apparatus further includes a data filtering module, configured to obtain a sensor identifier corresponding to the point cloud data; acquiring a corresponding filtering threshold according to the sensor identifier; and screening the point data based on the filtering threshold value to obtain the screened point data.
In one embodiment, the threshold generating module 406 is further configured to compare a vertical axis coordinate corresponding to a midpoint of the sub-point cloud with a segmentation threshold; and when the vertical axis coordinate is greater than or equal to the segmentation threshold, determining a point corresponding to the vertical axis coordinate as a non-ground point.
For specific limitations of the point cloud based segmentation processing device, reference may be made to the above limitations of the point cloud based segmentation processing method, which is not described herein again. The modules in the point cloud based segmentation processing device can be wholly or partially realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent of a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, an internal structure of which may be as shown in fig. 5. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer readable instructions, and a database. The internal memory provides an environment for the operating system and execution of computer-readable instructions in the non-volatile storage medium. The database of the computer device is used for storing point cloud-based segmentation processing data. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer readable instructions, when executed by a processor, implement a point cloud based segmentation processing method.
Those skilled in the art will appreciate that the architecture shown in fig. 5 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and one or more processors, the memory having stored therein computer-readable instructions, which, when executed by the processors, cause the one or more processors to perform the steps of the above-described method embodiments.
In one embodiment, one or more non-transitory computer-readable storage media storing computer-readable instructions are provided, which when executed by one or more processors, cause the one or more processors to perform the steps of the above-described method embodiments when executed.
In one embodiment, a vehicle is provided, and the vehicle may specifically include an autonomous vehicle, an electric vehicle, a bicycle, an aircraft, and the like, and the vehicle includes the above computer device, and may perform the steps in the above point cloud-based segmentation processing method embodiment.
The embodiments and implementation objects of the present invention are not limited to autonomous vehicles, electric vehicles, bicycles, aircrafts, robots, etc., but also include simulation devices, test equipment, etc. applied to these devices.
It will be understood by those of ordinary skill in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware associated with computer readable instructions, which can be stored in a non-volatile computer readable storage medium, and when executed, can include processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), rambus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent application shall be subject to the appended claims.
Claims (10)
1. A point cloud-based segmentation processing method comprises the following steps:
acquiring point cloud data, wherein the point cloud data comprises point data corresponding to a plurality of points, and the point data comprises three-dimensional coordinates corresponding to the points;
dividing the point cloud data into a plurality of sub-point clouds according to the point data, and mapping three-dimensional coordinates corresponding to the points to a polar coordinate system; generating a plurality of polar coordinate grids according to the polar coordinate system; counting the points in the polar coordinate grid to obtain sub-point clouds corresponding to the polar coordinate grid;
generating a segmentation threshold corresponding to the sub-point cloud based on point data corresponding to a plurality of points in the sub-point cloud, wherein the segmentation threshold corresponding to the sub-point cloud comprises the steps of obtaining vertical axis coordinates corresponding to the points from the point data, evaluating ground height coordinates according to the vertical axis coordinates, and calculating the sum of the ground height coordinates and a preset height threshold to obtain the segmentation threshold corresponding to the sub-point cloud; the method comprises the steps of evaluating a ground height coordinate according to a plurality of vertical axis coordinates, and sequencing a plurality of points according to the vertical axis coordinates to obtain a point sequence; acquiring points at preset positions in the point sequence and recording the points as points to be evaluated; calculating the average value of the vertical axis coordinates corresponding to the to-be-evaluated point based on the vertical axis coordinates corresponding to the to-be-evaluated point, and taking the average value as the ground height coordinates corresponding to the sub-point cloud;
traversing the plurality of points in the sub-point cloud, including comparing vertical axis coordinates corresponding to the points in the sub-point cloud with the segmentation threshold; when the vertical axis coordinate is larger than or equal to the segmentation threshold, determining a point corresponding to the vertical axis coordinate as a non-ground point; determining a point of which the vertical axis coordinate is smaller than the segmentation threshold value in the point data as a ground point; and
and counting the ground points in the sub-point clouds to obtain a ground point set corresponding to the point cloud data.
2. The method of claim 1, wherein the point cloud data is acquired from a laser sensor comprising a first sensor and a second sensor, the method further comprising:
acquiring a position relation between the second sensor and the first sensor; and
and mapping the point cloud data corresponding to the second sensor to a coordinate system corresponding to the first sensor according to the position relation.
3. The method of claim 1, wherein prior to said generating segmentation thresholds for said sub-point clouds based on point data corresponding to a plurality of said points in said sub-point clouds, said method further comprises:
acquiring a sensor identifier corresponding to the point cloud data;
acquiring a corresponding filtering threshold according to the sensor identifier; and
and screening the point data based on the filtering threshold value to obtain the screened point data.
4. A point cloud based segmentation processing apparatus comprising:
the system comprises a point cloud obtaining module, a point cloud obtaining module and a point cloud processing module, wherein the point cloud obtaining module is used for obtaining point cloud data, the point cloud data comprises point data corresponding to a plurality of points, and the point data comprises three-dimensional coordinates corresponding to the points;
the point cloud dividing module is used for dividing the point cloud data into a plurality of sub-point clouds according to the point data, and mapping the corresponding three-dimensional coordinates to a polar coordinate system; generating a plurality of polar coordinate grids according to the polar coordinate system; counting the points in the polar coordinate grid to obtain sub-point clouds corresponding to the polar coordinate grid;
a threshold generating module, configured to generate a segmentation threshold corresponding to the sub-point cloud based on point data corresponding to a plurality of points in the sub-point cloud, where the step of generating the segmentation threshold corresponding to the sub-point cloud includes obtaining vertical axis coordinates corresponding to the plurality of points from the point data, evaluating a ground height coordinate according to the plurality of vertical axis coordinates, and calculating a sum of the ground height coordinate and a preset height threshold to obtain the segmentation threshold corresponding to the sub-point cloud; the method comprises the steps of evaluating a ground height coordinate according to a plurality of vertical axis coordinates, and sequencing a plurality of points according to the vertical axis coordinates to obtain a point sequence; acquiring points at preset positions in the point sequence and recording the points as points to be evaluated; calculating the average value of the vertical axis coordinates corresponding to the to-be-evaluated point based on the vertical axis coordinates corresponding to the to-be-evaluated point, and taking the average value as the ground height coordinates corresponding to the sub-point cloud; traversing the plurality of points in the sub-point cloud, including comparing vertical axis coordinates corresponding to the points in the sub-point cloud with the segmentation threshold; when the vertical axis coordinate is larger than or equal to the segmentation threshold, determining a point corresponding to the vertical axis coordinate as a non-ground point; determining a point of which the vertical axis coordinate is smaller than the segmentation threshold value in the point data as a ground point; and
and the ground point counting module is used for counting ground points in the sub-point clouds to obtain a ground point set corresponding to the point cloud data.
5. A computer device comprising a memory and one or more processors, the memory having stored therein computer-readable instructions that, when executed by the one or more processors, cause the one or more processors to perform the steps of:
acquiring point cloud data, wherein the point cloud data comprises point data corresponding to a plurality of points, and the point data comprises three-dimensional coordinates corresponding to the points;
dividing the point cloud data into a plurality of sub-point clouds according to the point data, and mapping three-dimensional coordinates corresponding to the points to a polar coordinate system; generating a plurality of polar coordinate grids according to the polar coordinate system; counting the points in the polar coordinate grid to obtain sub-point clouds corresponding to the polar coordinate grid;
generating a segmentation threshold corresponding to the sub-point cloud based on point data corresponding to a plurality of points in the sub-point cloud, wherein the segmentation threshold corresponding to the sub-point cloud comprises the steps of obtaining vertical axis coordinates corresponding to the points from the point data, evaluating ground height coordinates according to the vertical axis coordinates, and calculating the sum of the ground height coordinates and a preset height threshold to obtain the segmentation threshold corresponding to the sub-point cloud; the method comprises the steps of evaluating a ground height coordinate according to a plurality of vertical axis coordinates, and sequencing a plurality of points according to the vertical axis coordinates to obtain a point sequence; acquiring points at preset positions in the point sequence and recording the points as points to be evaluated; calculating the average value of the vertical axis coordinates corresponding to the to-be-evaluated point based on the vertical axis coordinates corresponding to the to-be-evaluated point, and taking the average value as the ground height coordinates corresponding to the sub-point cloud;
traversing the plurality of points in the sub-point cloud, including comparing vertical axis coordinates corresponding to the points in the sub-point cloud with the segmentation threshold; when the vertical axis coordinate is larger than or equal to the segmentation threshold, determining a point corresponding to the vertical axis coordinate as a non-ground point; determining a point of which the vertical axis coordinate in the point data is smaller than the segmentation threshold value as a ground point; and
and counting the ground points in the sub-point clouds to obtain a ground point set corresponding to the point cloud data.
6. The computer device of claim 5, wherein the point cloud data is collected from a laser sensor, the laser sensor comprising a first sensor and a second sensor, the processor when executing the computer readable instructions further performing the steps of:
acquiring a position relation between the second sensor and the first sensor; and
and mapping the point cloud data corresponding to the second sensor to a coordinate system corresponding to the first sensor according to the position relation.
7. The computer device of claim 5, wherein the processor, when executing the computer readable instructions, further performs the steps of:
acquiring a sensor identifier corresponding to the point cloud data;
acquiring a corresponding filtering threshold according to the sensor identifier; and
and screening the point data based on the filtering threshold value to obtain the screened point data.
8. A non-transitory computer-readable storage medium storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of:
acquiring point cloud data, wherein the point cloud data comprises point data corresponding to a plurality of points, and the point data comprises three-dimensional coordinates corresponding to the points;
dividing the point cloud data into a plurality of sub-point clouds according to the point data, and mapping three-dimensional coordinates corresponding to the points to a polar coordinate system; generating a plurality of polar coordinate grids according to the polar coordinate system; counting the points in the polar coordinate grid to obtain sub-point clouds corresponding to the polar coordinate grid;
generating a segmentation threshold corresponding to the sub-point cloud based on point data corresponding to a plurality of points in the sub-point cloud, wherein the segmentation threshold corresponding to the sub-point cloud comprises the steps of obtaining vertical axis coordinates corresponding to the points from the point data, evaluating ground height coordinates according to the vertical axis coordinates, and calculating the sum of the ground height coordinates and a preset height threshold to obtain the segmentation threshold corresponding to the sub-point cloud; the method comprises the steps of evaluating a ground height coordinate according to a plurality of vertical axis coordinates, and sequencing a plurality of points according to the vertical axis coordinates to obtain a point sequence; acquiring points at preset positions in the point sequence and recording the points as points to be evaluated; calculating the average value of the vertical axis coordinates corresponding to the to-be-evaluated point based on the vertical axis coordinates corresponding to the to-be-evaluated point, and taking the average value as the ground height coordinate corresponding to the sub-point cloud;
traversing the plurality of points in the sub-point cloud, including comparing vertical axis coordinates corresponding to the points in the sub-point cloud with the segmentation threshold; when the vertical axis coordinate is larger than or equal to the segmentation threshold, determining a point corresponding to the vertical axis coordinate as a non-ground point; determining a point of which the vertical axis coordinate is smaller than the segmentation threshold value in the point data as a ground point; and
and counting the ground points in the sub-point clouds to obtain a ground point set corresponding to the point cloud data.
9. The storage medium of claim 8, wherein the point cloud data is acquired from a laser sensor comprising a first sensor and a second sensor, the computer readable instructions when executed by the processor further performing the steps of:
acquiring a position relation between the second sensor and the first sensor; and
and mapping the point cloud data corresponding to the second sensor to a coordinate system corresponding to the first sensor according to the position relation.
10. A vehicle comprising a memory and one or more processors, the memory having stored therein computer-readable instructions that, when executed by the one or more processors, cause the one or more processors to perform the method of point cloud based segmentation processing according to any one of claims 1-3.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2019/130170 WO2021134339A1 (en) | 2019-12-30 | 2019-12-30 | Point cloud-based segmentation processing method and apparatus, computer device, and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113366532A CN113366532A (en) | 2021-09-07 |
| CN113366532B true CN113366532B (en) | 2023-03-21 |
Family
ID=76686217
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201980037714.0A Active CN113366532B (en) | 2019-12-30 | 2019-12-30 | Point cloud based segmentation processing method and device, computer equipment and storage medium |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN113366532B (en) |
| WO (1) | WO2021134339A1 (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114119628B (en) * | 2021-10-25 | 2022-10-11 | 南京航空航天大学 | Point cloud accurate segmentation method based on feature template |
| CN114298581A (en) * | 2021-12-30 | 2022-04-08 | 广州极飞科技股份有限公司 | Quality evaluation model generation method, quality evaluation device, electronic device, and readable storage medium |
| CN115824237B (en) * | 2022-11-29 | 2023-09-26 | 重庆赛迪奇智人工智能科技有限公司 | Rail pavement recognition method and device |
| CN116071231B (en) * | 2022-12-16 | 2023-12-29 | 群滨智造科技(苏州)有限公司 | Method, device, equipment and medium for generating ink-dispensing process track of glasses frame |
| CN116413740B (en) * | 2023-06-09 | 2023-09-05 | 广汽埃安新能源汽车股份有限公司 | Laser radar point cloud ground detection method and device |
| CN116452403B (en) * | 2023-06-16 | 2023-09-01 | 瀚博半导体(上海)有限公司 | Point cloud data processing method and device, computer equipment and storage medium |
| CN117475002A (en) * | 2023-12-27 | 2024-01-30 | 青岛亿联建设集团股份有限公司 | Building inclination measuring method based on laser scanning technology |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2010200144A1 (en) * | 2010-01-14 | 2011-07-28 | The University Of Sydney | Extraction processes |
| CN104298998A (en) * | 2014-09-28 | 2015-01-21 | 北京理工大学 | 3D point cloud data processing method |
| CN106204705A (en) * | 2016-07-05 | 2016-12-07 | 长安大学 | A kind of 3D point cloud segmentation method based on multi-line laser radar |
| CN108873013A (en) * | 2018-06-27 | 2018-11-23 | 江苏大学 | A kind of road using multi-line laser radar can traffic areas acquisition methods |
| CN110197215A (en) * | 2019-05-22 | 2019-09-03 | 深圳市牧月科技有限公司 | A kind of ground perception point cloud semantic segmentation method of autonomous driving |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104143194B (en) * | 2014-08-20 | 2017-09-08 | 清华大学 | A kind of point cloud segmentation method and device |
-
2019
- 2019-12-30 WO PCT/CN2019/130170 patent/WO2021134339A1/en active Application Filing
- 2019-12-30 CN CN201980037714.0A patent/CN113366532B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2010200144A1 (en) * | 2010-01-14 | 2011-07-28 | The University Of Sydney | Extraction processes |
| CN104298998A (en) * | 2014-09-28 | 2015-01-21 | 北京理工大学 | 3D point cloud data processing method |
| CN106204705A (en) * | 2016-07-05 | 2016-12-07 | 长安大学 | A kind of 3D point cloud segmentation method based on multi-line laser radar |
| CN108873013A (en) * | 2018-06-27 | 2018-11-23 | 江苏大学 | A kind of road using multi-line laser radar can traffic areas acquisition methods |
| CN110197215A (en) * | 2019-05-22 | 2019-09-03 | 深圳市牧月科技有限公司 | A kind of ground perception point cloud semantic segmentation method of autonomous driving |
Non-Patent Citations (1)
| Title |
|---|
| Fast segmentation of 3D point clouds: A paradigm on LiDAR data for autonomous vehicle applications;Dimitris Zermas等;《2017 IEEE International Conference on Robotics and Automation (ICRA)》;20170603;第1-7页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113366532A (en) | 2021-09-07 |
| WO2021134339A1 (en) | 2021-07-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113366532B (en) | Point cloud based segmentation processing method and device, computer equipment and storage medium | |
| CN110458854B (en) | Road edge detection method and device | |
| CN110084116B (en) | Road surface detection method, road surface detection device, computer equipment and storage medium | |
| CN111160302A (en) | Obstacle information identification method and device based on automatic driving environment | |
| CN106886980B (en) | Point cloud density enhancement method based on three-dimensional laser radar target identification | |
| CN105404844A (en) | Road boundary detection method based on multi-line laser radar | |
| CN113424079A (en) | Obstacle detection method, obstacle detection device, computer device, and storage medium | |
| CN114981840A (en) | Ground segmentation method and device based on point cloud data and computer equipment | |
| CN115917559A (en) | Trajectory prediction method, apparatus, computer device and storage medium | |
| CN112149550A (en) | Automatic driving vehicle 3D target detection method based on multi-sensor fusion | |
| WO2022099530A1 (en) | Motion segmentation method and apparatus for point cloud data, computer device and storage medium | |
| CN113490965A (en) | Image tracking processing method and device, computer equipment and storage medium | |
| CN108074232B (en) | Voxel segmentation-based airborne LIDAR building detection method | |
| CN107796373B (en) | Distance measurement method based on monocular vision of front vehicle driven by lane plane geometric model | |
| CN114930401A (en) | Point cloud-based three-dimensional reconstruction method and device and computer equipment | |
| CN111144228B (en) | Obstacle identification method based on 3D point cloud data and computer equipment | |
| WO2021185121A1 (en) | Model generation method and apparatus, object detection method and apparatus, device, and storage medium | |
| WO2022133770A1 (en) | Method for generating point cloud normal vector, apparatus, computer device, and storage medium | |
| CN113008296B (en) | Method for detecting the environment of a vehicle by fusing sensor data on a point cloud plane and vehicle control unit | |
| CN114821526A (en) | Obstacle three-dimensional frame detection method based on 4D millimeter wave radar point cloud | |
| WO2022226831A1 (en) | Method and apparatus for detecting category-undefined obstacle, and computer device | |
| CN114170188A (en) | Target counting method and system for overlook image and storage medium | |
| CN114241448A (en) | Method and device for obtaining heading angle of obstacle, electronic equipment and vehicle | |
| WO2022141494A1 (en) | Three-dimensional point cloud clustering method and apparatus, computer device, and storage medium | |
| CN114005045A (en) | Rotating frame remote sensing target detection method based on lightweight deep neural network |
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 |