CN-119881936-B - Point cloud data processing method and device, electronic equipment and storage medium

CN119881936BCN 119881936 BCN119881936 BCN 119881936BCN-119881936-B

Abstract

The invention provides a point cloud data processing method, a device, electronic equipment and a storage medium, which comprise the steps of obtaining a point cloud data block in a current acquisition frame of a laser radar and measurement data of an inertial measurement unit, wherein the point cloud data block is obtained by dividing laser points acquired by the laser radar in advance, adjusting each point cloud data block to a frame head position of the current acquisition frame, generating rotation conversion quantity and translation conversion quantity of each point cloud data block by adopting the measurement data, determining the point cloud data block corresponding to the laser points, and carrying out the same motion compensation on the laser points by adopting the rotation conversion quantity and the translation conversion quantity of the point cloud data block to obtain compensated target point cloud data. According to the method, one frame of laser radar data is divided into a plurality of data blocks, the rotation conversion amount and the translation conversion amount of the point cloud data blocks are calculated, and the laser points on each point cloud data block directly execute the same rotation conversion and translation conversion, so that the efficient compensation for the motion distortion of the laser radar point cloud is realized.

Inventors

HUANG HONG
DENG HAOYUN
YANG KAI

Assignees

重庆长安科技有限责任公司

Dates

Publication Date: 20260508
Application Date: 20250113

Claims (9)

1. A method for processing point cloud data, the method comprising: Acquiring a point cloud data block in a current acquisition frame of the laser radar and measurement data of an inertial measurement unit, wherein the point cloud data block is obtained by dividing laser points acquired in the current acquisition frame of the laser radar according to the number of UDP packets in advance; The method comprises the steps of determining target measurement data of a current acquisition frame head moment, adjusting the point cloud data block to the frame head moment by using the target measurement data, acquiring ordered time stamps of the point cloud data block, determining measurement data of a first moment and a second moment by using the ordered time stamps, and obtaining rotation quaternion variation and translation variation of the point cloud data block from the first moment to the second moment, wherein the ordered time stamp moment is between the first moment and the second moment, and the rotation quaternion variation and translation variation of the point cloud data block are generated by using the rotation quaternion variation and the translation variation of the first moment to the second moment; And determining a point cloud data block corresponding to the laser point, and performing the same motion compensation on the laser point by adopting the rotation conversion amount and the translation conversion amount of the point cloud data block to obtain compensated target point cloud data.
2. The method of claim 1, wherein the acquiring the point cloud data block of the current acquisition frame of the laser radar and the measurement data of the inertial measurement unit, wherein the point cloud data block is obtained by dividing laser points acquired by the laser radar in advance, comprises: performing time synchronization on a laser radar and an inertial measurement unit in advance; Acquiring laser points acquired by a laser radar in a current acquisition frame, and dividing the laser points into data blocks by adopting the number of preset packets to obtain point cloud data blocks of the current acquisition frame, wherein the number of the point cloud data blocks is the same as the number of the preset packets; And acquiring a point cloud data block of the laser radar after time synchronization and measurement data of an inertial measurement unit.
3. The method of claim 1, wherein determining target measurement data for a frame start time of a current acquisition frame, and adjusting the point cloud data block to the frame start time using the target measurement data, comprises: Acquiring first measurement data before the frame head moment of the current acquisition frame and second measurement data after the frame tail moment of the current acquisition frame; Determining target measurement data at the frame head moment by adopting first measurement data and second measurement data, wherein the target measurement data comprises a rotation quaternion and a triaxial coordinate; And adjusting the point cloud data block to the frame head moment by adopting the rotation quaternion and the triaxial coordinate.
4. The method of claim 1, wherein generating the rotational and translational amounts of the point cloud data block using the rotational quaternion amounts of change and the translational amounts of change from the first time to the second time comprises: Obtaining interpolation parameters of ordered time stamps from the first moment to the point cloud data block by adopting the rotation quaternion variable quantity and the translation variable quantity from the first moment to the second moment; and generating the rotation conversion quantity and the translation conversion quantity of the point cloud data block by adopting the rotation quaternion variation quantity, the translation variation quantity and the interpolation parameter.
5. The method according to claim 1, wherein the determining the point cloud data block corresponding to the laser point, and performing motion compensation on the laser point by using the rotation conversion amount and the translation conversion amount of the point cloud data block, to obtain compensated target point cloud data, includes: determining the serial number of the laser points and the number of the laser points in the point cloud data block; Determining a point cloud data block to which the laser point belongs according to the serial number of the laser point and the number of the laser points in the point cloud data block; and carrying out rotary transformation on the laser point by adopting the rotary transformation amount of the point cloud data block, and carrying out translational transformation on the laser point after rotary transformation by adopting the translational transformation amount of the point cloud data block to obtain compensated target point cloud data.
6. A point cloud data processing apparatus, the apparatus comprising: The acquisition data module is used for acquiring a point cloud data block in a current acquisition frame of the laser radar and measurement data of the inertial measurement unit, wherein the point cloud data block is obtained by dividing laser points acquired in the current acquisition frame of the laser radar according to the number of UDP packets in advance; The generation compensation quantity module is used for adjusting each point cloud data block to the frame head position of a current acquisition frame, generating rotation conversion quantity and translation conversion quantity of each point cloud data block by adopting the measurement data, and comprises an adjustment sub-module, a first determination sub-module and a generation sub-module, wherein the adjustment sub-module is used for determining target measurement data of the current acquisition frame head moment, adjusting the point cloud data block to the frame head moment by adopting the target measurement data, the first determination sub-module is used for acquiring ordered time stamps of the point cloud data block, determining measurement data of a first moment and a second moment by adopting the ordered time stamps, and obtaining rotation quaternion change quantity and translation change quantity from the first moment to the second moment, the ordered time stamp moment is between the first moment and the second moment, and the generation sub-module is used for generating rotation conversion quantity and translation change quantity of the point cloud data block by adopting the first moment to the second moment; And the data compensation module is used for determining a point cloud data block corresponding to the laser point, and carrying out the same motion compensation on the laser point by adopting the rotation conversion amount and the translation conversion amount of the point cloud data block to obtain compensated target point cloud data.
7. The apparatus of claim 6, wherein the means for obtaining data comprises: the synchronization sub-module is used for carrying out time synchronization on the laser radar and the inertial measurement unit in advance; Dividing the sub-module, which is used for obtaining laser points collected by the laser radar in the current collection frame, and dividing the data blocks of the laser points by adopting the number of preset message packets to obtain point cloud data blocks of the current collection frame, wherein the number of the point cloud data blocks is the same as the number of the preset message packets; The acquisition sub-module is used for acquiring the point cloud data block of the laser radar after time synchronization and the measurement data of the inertial measurement unit.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the computer program when executed by the processor implements the point cloud data processing method of any of claims 1 to 5.
9. A readable storage medium, characterized in that the readable storage medium has stored thereon a computer program which, when executed by a processor, implements the point cloud data processing method according to any of claims 1 to 5.

Description

Point cloud data processing method and device, electronic equipment and storage medium Technical Field The embodiment of the invention relates to the technical field of automatic driving, in particular to a point cloud data processing method, a device, electronic equipment and a storage medium. Background Lidar is a high-precision active sensor that is increasingly being used in high-order autopilot schemes. When the laser radar is mounted on an automatic driving vehicle, as the principle of scanning imaging of the laser radar sensor is a slow exposure process and the vehicle moves relative to the surrounding environment, the coordinate system of the laser radar sensor can be changed along with the movement of the vehicle in the same frame data scanning process, so that the movement distortion of point cloud data is generated, and the three-dimensional environment around the automatic driving vehicle at a certain moment can not be truly reflected. In order to obtain accurate point cloud data, motion distortion compensation is required, and the conventional TOF (Time-of-Flight) laser radar point cloud distortion compensation method is to perform corresponding transformation on each point cloud data by estimating the motion of a laser radar in a scanning process so as to eliminate distortion caused by the motion. Disclosure of Invention The invention aims to provide a point cloud data processing method, solve the problems of large calculation amount and low efficiency of point-by-point linear interpolation processing of a large amount of point cloud data, realize efficient point cloud data processing, provide a point cloud data processing device, provide electronic equipment and provide a storage medium. In order to achieve the above purpose, the technical scheme adopted by the invention is as follows: an embodiment of the present invention provides a method for processing point cloud data, where the method includes: Acquiring a point cloud data block in a current acquisition frame of the laser radar and measurement data of an inertial measurement unit, wherein the point cloud data block is obtained by dividing laser points acquired by the laser radar in advance; Adjusting each point cloud data block to the frame head position of the current acquisition frame, and generating the rotation conversion quantity and the translation conversion quantity of each point cloud data block by adopting the measurement data; And determining a point cloud data block corresponding to the laser point, and performing the same motion compensation on the laser point by adopting the rotation conversion amount and the translation conversion amount of the point cloud data block to obtain compensated target point cloud data. Optionally, the acquiring the point cloud data block of the current acquisition frame of the laser radar and the measurement data of the inertial measurement unit, wherein the point cloud data block is obtained by dividing the laser points acquired by the laser radar in advance, and comprises the following steps: performing time synchronization on a laser radar and an inertial measurement unit in advance; Acquiring laser points acquired by a laser radar in a current acquisition frame, and dividing the laser points into data blocks by adopting the number of preset packets to obtain point cloud data blocks of the current acquisition frame, wherein the number of the point cloud data blocks is the same as the number of the preset packets; And acquiring a point cloud data block of the laser radar after time synchronization and measurement data of an inertial measurement unit. Optionally, the adjusting the cloud data blocks of each point to the frame head position of the current acquisition frame, generating the rotation conversion amount and the translation conversion amount of the cloud data blocks of each point by using the measurement data includes: Determining target measurement data of the frame head moment of the current acquisition frame, and adjusting the point cloud data block to the frame head moment by adopting the target measurement data; Acquiring an ordered time stamp of a point cloud data block, and determining measurement data of a first moment and a second moment by adopting the ordered time stamp to obtain a rotation quaternion variation and a translation variation from the first moment to the second moment; and generating the rotation conversion amount and the translation conversion amount of the point cloud data block by adopting the rotation quaternion change amount and the translation change amount from the first moment to the second moment. Optionally, determining the target measurement data of the frame head time of the current acquisition frame, and adjusting the point cloud data block to the frame head time by using the target measurement data includes: Acquiring first measurement data before the frame head moment of the current acquisition frame and second measurement data after the frame ta