Sensor calibration method and device, computer device, medium, and vehicle
Embodiments of the present disclosure provide a sensor calibration method, a sensor calibration device, a computer device, a storage medium, and a vehicle. The method includes: detecting surrounding objects in a travelling process of a vehicle; recognizing a static object from the surrounding objects; performing feature extraction on the static object by a camera and a lidar, respectively; and calibrating an extrinsic parameter of the camera and the lidar based on the extracted feature.
1. A sensor calibration method, comprising:
detecting surrounding objects in a travelling process of a vehicle;
recognizing a static object from the surrounding objects;
performing feature extraction on the static object by a camera and a lidar, respectively; and
calibrating an extrinsic parameter of the camera and the lidar based on the extracted feature,
wherein performing feature extraction on the static object by the camera and the lidar and calibrating the extrinsic parameter of the camera and the lidar based on the extracted feature comprises:
performing feature extraction on an edge of the static object by the camera and the lidar respectively to obtain a plurality of feature points on the edge, and obtaining a first coordinate of each of the plurality of feature points in a camera imaging plane and a second coordinate of each of the plurality of feature points in a lidar coordinate system;
for each of the plurality of feature points, performing an alignment operation between a coordinate of the second coordinate projected on the camera imaging plane and the first coordinate by an iterative method; and
determining a rotation matrix and a translation vector between the camera and the lidar when the alignment is achieved.
2. The sensor calibration method according to claim 1 , wherein detecting the surrounding objects comprises:
performing obstacle detection by the camera and the lidar, respectively;
performing verification on an obstacle detection result of the camera and an obstacle detection result of the lidar, to obtain a final obstacle detection result; and
determining the surrounding objects based on the final obstacle detection result.
3. The sensor calibration method according to claim 1 , wherein recognizing the static object from the surrounding objects comprises:
recognizing the static object from the surrounding objects based on a pre-trained recognition model.
4. The sensor calibration method according to claim 1 , wherein the edge comprises one or more of a straight line and a circular ring.
5. The sensor calibration method according to claim 1 , wherein the static object comprises one or more of a car, a house, a tree, and a utility pole.
6. A sensor calibration device, comprising:
one or more processors;
a storage device, configured to store one or more programs,
wherein when the one or more programs are executed by the one or more processors, the one or more processors are caused to:
detect surrounding objects in a travelling process of a vehicle;
recognize a static object from the surrounding objects; and
perform feature extraction on the static object by a camera and a lidar respectively, and calibrate an extrinsic parameter of the camera and the lidar based on the extracted feature;
wherein in performing feature extraction on the static object by the camera and the lidar and calibrating the extrinsic parameter of the camera and the lidar based on the extracted feature, the one or more processors are configured to:
perform feature extraction on an edge of the static object by the camera and the lidar respectively to obtain a plurality of feature points on the edge, and obtain a first coordinate of each of the plurality of feature points in a camera imaging plane and a second coordinate of each of the plurality of feature points in a lidar coordinate system; and
for each of the plurality of feature points, perform an alignment operation between a coordinate of the second coordinate projected on the camera imaging plane and the first coordinate by an iterative method, and determine a rotation matrix and a translation vector between the camera and the lidar when the alignment is achieved.
7. The sensor calibration device according to claim 6 , wherein in detecting the surrounding objects, the one or more processors are configured to:
perform obstacle detection by the camera and the lidar respectively;
perform verification on an obstacle detection result of the camera and an obstacle detection result of the lidar, to obtain a final obstacle detection result; and
determine the surrounding objects based on the final obstacle detection result.
8. The sensor calibration device according to claim 6 , wherein in recognizing the static object from the surrounding objects, the one or more processors are configured to:
recognize the static object from the surrounding objects based on a pre-trained recognition model.
9. The sensor calibration device according to claim 6 , wherein the edge comprises one or more of a straight line and a circular ring.
10. The sensor calibration device according to claim 6 , wherein the static object comprises one or more of a car, a house, a tree, and a utility pole.
11. A non-transitory computer storage medium having stored thereon a computer program that, when executed by a processor, causes a sensor calibration method to be implemented, the method comprising:
detecting surrounding objects in a travelling process of a vehicle;
recognizing a static object from the surrounding objects;
performing feature extraction on the static object by a camera and a lidar, respectively; and
calibrating an extrinsic parameter of the camera and the lidar based on the extracted feature,
wherein performing feature extraction on the static object by the camera and the lidar and calibrating the extrinsic parameter of the camera and the lidar based on the extracted feature comprises:
performing feature extraction on an edge of the static object by the camera and the lidar respectively to obtain a plurality of feature points on the edge, and obtaining a first coordinate of each of the plurality of feature points in a camera imaging plane and a second coordinate of each of the plurality of feature points in a lidar coordinate system;
for each of the plurality of feature points, performing an alignment operation between a coordinate of the second coordinate projected on the camera imaging plane and the first coordinate by an iterative method; and
determining a rotation matrix and a translation vector between the camera and the lidar when the alignment is achieved.
12. The non-transitory computer storage medium according to claim 11 , wherein detecting the surrounding objects comprises:
performing obstacle detection by the camera and the lidar, respectively;
performing verification on an obstacle detection result of the camera and an obstacle detection result of the lidar, to obtain a final obstacle detection result; and
determining the surrounding objects based on the final obstacle detection result.
13. The non-transitory computer storage medium according to claim 11 , wherein recognizing the static object from the surrounding objects comprises:
recognizing the static object from the surrounding objects based on a pre-trained recognition model.
14. The non-transitory computer storage medium according to claim 11 , wherein the edge comprises one or more of a straight line and a circular ring.
15. The non-transitory computer storage medium according to claim 11 , wherein the static object comprises one or more of a car, a house, a tree, and a utility pole.