pkg_camera_calibration

中文版本

CeleX5-MIPI Intrinsics Calibration Demo

Calibrate CeleX5-MIPI camera intrinsics with camera-calibration.

• Calibration principle
• Running
• Example results

We use the blinking chessboard calibration board (generated with pattern-generator in this repository) to calibrate the CeleX™ CeleX5-MIPI Event-based Camera.

Calibration principle

By using the blinking chessboard, the CeleX5-MIPI camera can generate a chessboard like event frame, and then the traditional Camera Intrinsics calibration method can be used. Here camera calibration calibration tool integrated in ROS is used for calibration, and Zhang Zhengyou-Calibration method is used internally. The calibration results are saved with .yaml file, .txt file and image frame for final calibration results.

Running

1. First install camera-calibration.

   sudo apt install ros-kinetic-camera-calibration

2. Run celex5_ros_node:

   roslaunch celex5_ros celex5_ros_node.launch

   Open rqt_reconfigure panel, under Event Off-Pixel Timestamp Mode sensor mode, only keep /celex5_mipi/display/binary_img Image topic publishing messages, close other Image topics to reduce running consumption. Keep /celex5_mipi/display/binary_img publishing frequency at 20 fps, too high can cause a delay in the calibration procedure.

3. Launch camera-calibration. For the detailed meaning and more usage of each parameter, please go to the official document of camera-calibration in the above link.

   rosrun camera_calibration cameracalibrator.py --size 8x6 --square 0.0212 image:=/celex5_mipi/display/binary_img

   The meaning of parameters:

   • --size: The number of inner points (that is, the number of squares in the outer circle with the longest side out - 1), width*height. If this parameter is not set correctly, the calibration procedure will be found to be unresponsive in the following calibration steps.
   • --square: the physical length of each side of the squares in the calibration board. The unit is m.
   • image:=: Image topic name.

   After successful operation, the following windows will appear:

   

   Move the blinking chessboard screen, and make the X, Y, Size and Skew progress bars in the upper right corner full and green as much as possible.

   Reference the official document of camera-calibration:

   • checkerboard on the camera's left, right, top and bottom of field of view
   • X bar - left/right in field of view
   • Y bar - top/bottom in field of view
   • Size bar - toward/away and tilt from the camera
   • checkerboard filling the whole field of view
   • checkerboard tilted to the left, right, top and bottom (Skew)

Tips: In the process of moving, the blinking chessboard pattern can be suspended, so as to avoid corner detection when the dragging is serious. After fixing the position, continue to make the checkerboard blinking to obtain a high-quality Event frame

When you see the CALIBRATE button turns green, click calculate to get the internal parameter calibration result. Then click the SAVE button to save the calibration results. It will prompt you to save the path, which is a compressed package. After decompression, the image frame used in the calibration process and calibration result file (one .yaml file and one .txt file) are stored.

Example results

This is the whole calibration output:

ost.yaml:

image_width: 1280
image_height: 800
camera_name: narrow_stereo
camera_matrix:
  rows: 3
  cols: 3
  data: [1687.514919, 0.000000, 676.131784, 0.000000, 1694.799737, 417.348889, 0.000000, 0.000000, 1.000000]
distortion_model: plumb_bob
distortion_coefficients:
  rows: 1
  cols: 5
  data: [-0.089225, -0.433622, 0.003022, -0.001306, 0.000000]
rectification_matrix:
  rows: 3
cols: 3
  data: [1.000000, 0.000000, 0.000000, 0.000000, 1.000000, 0.000000, 0.000000, 0.000000, 1.000000]
projection_matrix:
  rows: 3
cols: 4
  data: [1647.456299, 0.000000, 677.701102, 0.000000, 0.000000, 1683.463867, 418.507924, 0.000000, 0.000000, 0.000000, 1.000000, 0.000000]

The sample image frame used in the calibration: