Today, we are learning the camera calibration, according to
https://blog.csdn.net/heyijia0327/article/details/43538695
1.
According to my previous blog, run their own camera-SLAM2 ORB , proceed to 1.7 (the camera is turned on).
2. Check usb_cam package announcement:
rostopic list
To see if there are messages, and remember the name
/ usb_cam / camera_info
/ usb_cam / image_raw
3. Start calibration procedures:
rosrun camera_calibration cameracalibrator.py --size 8x5 --square 0.03 image:=/usb_cam_node/image_raw camera:=/usb_cam_node
Wherein the parameter is the letter size 11x8 noted x, he is the number of corners inside the checkerboard, as shown in FIG. square checkerboard square side length of 0.03m, note just the beginning calibration, CALIBRATE button is grayed out.
4. The calibration procedure using the method:
In order to obtain a good result of the calibration, the calibration plate should be such that as much as possible in various places in the camera's field of view:
Such as calibration plate on the horizon to the left, right, top and bottom, both calibration plate inclined, there are levels.
界面中的x:表示标定板在视野中的左右位置。
y:表示标定板在视野中的上下位置。
size:标定板在占视野的尺寸大小,也可以理解为标定板离摄像头的远近。
skew:标定板在视野中的倾斜位置。
And occupy the majority of the board but also the field of view of the image, as shown below:
pressing the above-described embodiment constantly moving the calibration plate, until the CALIBRATE button is lit, click the button will be calibrated. Calibration process will last twelve minutes, and calibration interface will be grayed out and can not operate, you can be patient.
After calibration is complete, you will see a window as shown below, the window image is a result of the calibration, to correct the distortion.
Also, the parameters of the camera will appear in the terminal window calibration program, as follows:
If the calibration satisfactory results, click on the COMMIT button to save the results to the default folder, you can see the following information in usb_cam window, the description has been saved calibration results when the appropriate folder next time you start usb_cam node will automatically call.
At last
Supplement the original blog does not have, this storage path can not be found, so I open the file with the command yaml
xdg-open head-camera.yaml