Method and apparatus for annotating point cloud data

1. A method for annotating point cloud data, comprising:

collecting data in a given scenario by using a laser radar and a non-laser radar sensor to respectively obtain point cloud data and sensor data;

segmenting and tracking the point cloud data to obtain point cloud segmentation and tracking results;

recognizing and tracking feature objects in the sensor data to obtain feature object recognition and tracking results;

matching the point cloud segmentation and tracking results with the feature object recognition and tracking results, to obtain confidence levels of the point cloud segmentation and tracking results;

acquiring a point cloud segmentation and tracking result having a confidence level greater than a confidence level threshold, for checking by a user;

acquiring the checked point cloud segmentation and tracking result in response to a user operation on the point cloud segmentation and tracking result having the confidence level greater than the confidence level threshold; and

determining the checked point cloud segmentation and tracking result as a point cloud annotation result,

wherein said collecting data in the given scenario by using the laser radar and the non- laser radar sensor to respectively obtain the point cloud data and the sensor data comprises:

collecting the data in the scenario by using a camera, to obtain a video, and

wherein said matching the point cloud segmentation and tracking results with the feature object recognition and tracking results, to obtain the confidence levels of the point cloud segmentation and tracking results comprises:

(i) determining whether a first object and a first movement trajectory of the first object in the point cloud segmentation and tracking results coincide with a second object and a second movement trajectory of the second object in the feature object recognition and tracking results obtained for the video; and

(ii) determining a coincidence degree of the first object and the first movement trajectory of the first object in the point cloud segmentation and tracking results with the second object and the second movement trajectory of the second object in the feature object recognition and tracking results obtained for the video; and outputting the determined coincidence degree as the confidence level.

2. The method according to claim 1 , wherein the recognizing and tracking feature objects in the sensor data to obtain feature object recognition and tracking results comprises:

executing a vehicle registration plate recognizing and tracking algorithm and/or a human face recognizing and tracking algorithm on video content of the video.

3. The method according to claim 1 , wherein the collecting data in a given scenario by using a laser radar and a non-laser radar sensor to respectively obtain point cloud data and sensor data comprises:

collecting data by using a laser radar and a sensor placed at the same heights as those mounted on a driverless vehicle.

4. The method according to claim 1 , wherein the collecting data in a given scenario by using a laser radar and a non-laser radar sensor to respectively obtain point cloud data and sensor data comprises:

collecting data by using a laser radar and a sensor mounted and calibrated on a driverless vehicle.

5. An apparatus for annotating point cloud data, comprising:

at least one processor; and

a memory storing instructions, the instructions when executed by the at least one processor, cause the at least one processor to perform operations, the operations comprising:

collecting data in a given scenario by using a laser radar and a non-laser radar sensor to respectively obtain point cloud data and sensor data;

segmenting and tracking the point cloud data to obtain point cloud segmentation and tracking results;

recognizing and tracking feature objects in the sensor data to obtain feature object recognition and tracking results;

matching the point cloud segmentation and tracking results with the feature object recognition and tracking results, to obtain confidence levels of the point cloud segmentation and tracking results;

acquiring a point cloud segmentation and tracking result having a confidence level greater than a confidence level threshold, for checking by a user;

acquiring the checked point cloud segmentation and tracking result in response to a user operation on the point cloud segmentation and tracking result having the confidence level greater than the confidence level threshold; and

determining the checked point cloud segmentation and tracking result as a point cloud annotation result,

wherein said collecting data in the given scenario by using the laser radar and the non-laser radar sensor to respectively obtain the point cloud data and the sensor data comprises:

collecting the data in the scenario by using a camera, to obtain a video, and

wherein said matching the point cloud segmentation and tracking results with the feature object recognition and tracking results, to obtain the confidence levels of the point cloud segmentation and tracking results comprises:

(i) determining whether a first object and a first movement trajectory of the first object in the point cloud segmentation and tracking results coincide with a second object and a second movement trajectory of the second object in the feature object recognition and tracking results obtained for the video; and

(ii) determining a coincidence degree of the first object and the first movement trajectory of the first object in the point cloud segmentation and tracking results with the second object and the second movement trajectory of the second object in the feature object recognition and tracking results obtained for the video; and outputting the determined coincidence degree as the confidence level.

6. The apparatus according to claim 5 , wherein the recognizing and tracking feature objects in the sensor data to obtain feature object recognition and tracking results comprises:

executing a vehicle registration plate recognizing and tracking algorithm and/or a human face recognizing and tracking algorithm on video content of the video.

7. The apparatus according to claim 5 , wherein the collecting data in a given scenario by using a laser radar and a non-laser radar sensor to respectively obtain point cloud data and sensor data comprises:

collecting data by using a laser radar and a sensor placed at the same heights as those mounted on a driverless vehicle.

8. The apparatus according to claim 5 , wherein the collecting data in a given scenario by using a laser radar and a non-laser radar sensor to respectively obtain point cloud data and sensor data comprises:

collecting data by using a laser radar and a sensor mounted and calibrated on a driverless vehicle.

9. A non-transitory computer storage medium storing a computer program, which when executed by a processor, cause the processor to perform operations, the operations comprising:

collecting data in a given scenario by using a laser radar and a non-laser radar sensor to respectively obtain point cloud data and sensor data;

segmenting and tracking the point cloud data to obtain point cloud segmentation and tracking results;

recognizing and tracking feature objects in the sensor data to obtain feature object recognition and tracking results;

matching the point cloud segmentation and tracking results by using the feature object recognition and tracking results, to obtain confidence levels of the point cloud segmentation and tracking results;

acquiring a point cloud segmentation and tracking result having a confidence level greater than a confidence level threshold, for checking by a user;

acquiring the checked point cloud segmentation and tracking result in response to a user operation on the point cloud segmentation and tracking result having the confidence level greater than the confidence level threshold; and

determining the checked point cloud segmentation and tracking result as a point cloud annotation result,

wherein said collecting data in the given scenario by using the laser radar and the non-laser radar sensor to respectively obtain the point cloud data and the sensor data comprises:

collecting the data in the scenario by using a camera, to obtain a video, and

wherein said matching the point cloud segmentation and tracking results with the feature object recognition and tracking results, to obtain the confidence levels of the point cloud segmentation and tracking results comprises:

(i) determining whether a first object and a first movement trajectory of the first object in the point cloud segmentation and tracking results coincide with a second object and a second movement trajectory of the second object in the feature object recognition and tracking results obtained for the video; and

(ii) determining a coincidence degree of the first object and the first movement trajectory of the first object in the point cloud segmentation and tracking results with the second object and the second movement trajectory of the second object in the feature object recognition and tracking results obtained for the video, and outputting the determined coincidence degree as the confidence level.

10. The non-transitory computer storage medium according to claim 9 , wherein the recognizing and tracking feature objects in the sensor data to obtain feature object recognition and tracking results comprises:

executing a vehicle registration plate recognizing and tracking algorithm and/or a human face recognizing and tracking algorithm on video content of the video.

11. The non-transitory computer storage medium according to claim 9 , wherein the collecting data in a given scenario by using a laser radar and a non-laser radar sensor to respectively obtain point cloud data and sensor data comprises:

collecting data by using a laser radar and a sensor placed at the same heights as those mounted on a driverless vehicle.

12. The non-transitory computer storage medium according to claim 9 , wherein the collecting data in a given scenario by using a laser radar and a non-laser radar sensor to respectively obtain point cloud data and sensor data comprises:

collecting data by using a laser radar and a sensor mounted and calibrated on a driverless vehicle.