Data Collection for Vehicle-3D Lidar calibration

PreviousOverlapping Camera Calibration (Multiple-Targets)NextData Collection for Vehicle-2D Lidar calibration

Last updated 2 years ago

Data Collection for Vehicle-3D Lidar calibration

Calibration Target:

Calibration target is a plane board. A bigger board can give a more accurate calibration result. The minimum board size is proportional to the vehicle size. For a small vehicle such as a car, at least 1 m in length and width is recommended. A large rolling white board or a large wardrobe cardboard box can be a good calibration target.

Example for rolling whiteboard:

Example for large cardboard box:

Data Collection:

Ensure the ground is flat for the vehicle and the locations of the boards.
Mark the vehicle boundary with tape on the floor. For a regular rectangular vehicle, please ensure that the front and rear boundary lines make right angles with the left and right boundary lines. You can use a large carpenter square to ensure the right angles.
Place a calibration target in front of the car in such a way that the plane of the board is parallel to the front boundary line of the car. Please also ensure that the board is perpendicular to the ground by using a carpenter square, level, or plumb bob. Take a LiDAR frame. Wait for 10 seconds (Don't move/rotate your car/robot/sensors). Stop recording. You must have a recording of lidar data for 10 seconds. Extract one frame of lidar data captured 5 seconds after recording has started (e.g. if you start recording at 3:00:00, you stop recording at 3:00:10. We need a frame captured at 3:00:05) and save them. (Dataset front_1) . Also note down the distance of the calibration target from the front vehicle boundary.
Similarly place the calibration target in the left of the car in such a way that the plane of the board is parallel to the left boundary line of the car and perpendicular to the ground. Collect a single frame like mentioned above. Also, note down the distance of the calibration target from the left vehicle boundary.
Similarly place the calibration target in the right of the car in such a way that the plane of the board is parallel to the right boundary line of the car and perpendicular to the ground. Collect a single frame like mentioned above. Also, note down the distance of the calibration target from the right vehicle boundary.
Similarly place the calibration target in the rear of the car in such a way that the plane of the board is parallel to the rear boundary of the car and perpendicular to the ground. Collect a single frame like mentioned above. Rear side data collection is optional. Also, note down the distance of the calibration target from the rear boundary of the vehicle.

Images taken while collecting data:

Tape marking the vehicle boundary

Image taken while collecting LiDAR frame after placing the board parallel to the front-side of vehicle:

Image taken while collecting LiDAR frame after placing the board parallel to the rear-side of vehicle:

Image taken while collecting LiDAR frame after placing the board parallel to the left-side of vehicle:

Image taken while collecting LiDAR frame after placing the board parallel to the right-side of vehicle:

Input configuration:

This calibration expects some input configuration for the vehicle. Based on which the extrinsic parameters optimized further.

Wheelbase: Distance between left/right front wheel center and left/right rear wheel center in meters. If asked for a Wheelbase without left/right, measure either the left or right wheelbase. If the wheelbase is the same on both left and right, the right wheelbase is optional.
Track: Distance between front/rear left wheel center and front/rear right wheel center in meters. If asked for just Track without front/rear, measure either the front or rear track. For a rectangular vehicle, this is also the distance between the left and right vehicle boundary lines.
Overhang: Distance between center of the front/rear wheel to the front/rear bumper of the vehicle in meters.

Note: For a rectangular vehicle, the distance between the front and rear vehicle boundary lines must be equal to the sum of wheelbase, front overhang, and rear overhang.

PreviousOverlapping Camera Calibration (Multiple-Targets)NextData Collection for Vehicle-2D Lidar calibration

Last updated 2 years ago

Calibration Target:

Example for rolling whiteboard:

Example for large cardboard box:

Data Collection:

Ensure the ground is flat for the vehicle and the locations of the boards.
Mark the vehicle boundary with tape on the floor. For a regular rectangular vehicle, please ensure that the front and rear boundary lines make right angles with the left and right boundary lines. You can use a large carpenter square to ensure the right angles.
Place a calibration target in front of the car in such a way that the plane of the board is parallel to the front boundary line of the car. Please also ensure that the board is perpendicular to the ground by using a carpenter square, level, or plumb bob. Take a LiDAR frame. Wait for 10 seconds (Don't move/rotate your car/robot/sensors). Stop recording. You must have a recording of lidar data for 10 seconds. Extract one frame of lidar data captured 5 seconds after recording has started (e.g. if you start recording at 3:00:00, you stop recording at 3:00:10. We need a frame captured at 3:00:05) and save them. (Dataset front_1) . Also note down the distance of the calibration target from the front vehicle boundary.
Similarly place the calibration target in the left of the car in such a way that the plane of the board is parallel to the left boundary line of the car and perpendicular to the ground. Collect a single frame like mentioned above. Also, note down the distance of the calibration target from the left vehicle boundary.
Similarly place the calibration target in the right of the car in such a way that the plane of the board is parallel to the right boundary line of the car and perpendicular to the ground. Collect a single frame like mentioned above. Also, note down the distance of the calibration target from the right vehicle boundary.
Similarly place the calibration target in the rear of the car in such a way that the plane of the board is parallel to the rear boundary of the car and perpendicular to the ground. Collect a single frame like mentioned above. Rear side data collection is optional. Also, note down the distance of the calibration target from the rear boundary of the vehicle.

Images taken while collecting data:

Tape marking the vehicle boundary

Image taken while collecting LiDAR frame after placing the board parallel to the front-side of vehicle:

Image taken while collecting LiDAR frame after placing the board parallel to the rear-side of vehicle:

Image taken while collecting LiDAR frame after placing the board parallel to the left-side of vehicle:

Image taken while collecting LiDAR frame after placing the board parallel to the right-side of vehicle:

Input configuration:

This calibration expects some input configuration for the vehicle. Based on which the extrinsic parameters optimized further.

Wheelbase: Distance between left/right front wheel center and left/right rear wheel center in meters. If asked for a Wheelbase without left/right, measure either the left or right wheelbase. If the wheelbase is the same on both left and right, the right wheelbase is optional.
Track: Distance between front/rear left wheel center and front/rear right wheel center in meters. If asked for just Track without front/rear, measure either the front or rear track. For a rectangular vehicle, this is also the distance between the left and right vehicle boundary lines.
Overhang: Distance between center of the front/rear wheel to the front/rear bumper of the vehicle in meters.

Note: For a rectangular vehicle, the distance between the front and rear vehicle boundary lines must be equal to the sum of wheelbase, front overhang, and rear overhang.