Shulou(Shulou.com)06/02 Report--

This article mainly introduces "what should be paid attention to when using kalibr calibration". In daily operation, I believe that many people have doubts about what they should pay attention to when using kalibr calibration. The editor consulted all kinds of data and sorted out simple and easy-to-use methods of operation. I hope it will be helpful to answer the doubts of "what should be paid attention to when using kalibr calibration". Next, please follow the editor to study!

1. The choice of model

The main purpose of this paper is to summarize the problems of model selection in using kalibr in various issue, as well as the advantages and disadvantages of various models.

Pinhole-radtan: the most common pinhole model + Brownian distortion model, which is suitable for most cameras with an angle less than 120, in which the distortion parameters include radial distortion k _ 1 ~ 2 and tangential distortion p _ 1 ~ p _ 2; if the distortion of the camera is not very serious, this model is basically fine; for example, my camera with a DOV of 150 can also use this and the effect of eliminating distortion is very good.

Pinhole-equi: pinhole model + isometric distortion model, which is the type that KB model needs to choose. This model also has a wide range of use, and most fisheye lenses can also be used. Note that the distortion parameter of the 8-parameter KB model is k _ 1 ~ k _ 2 ~ k _ 3 ~ ~ k _ 4, although it is also four numbers, but unlike the previous model, there are only radial distortion parameters, but no tangential distortion tangential distortion. The corresponding formula for projection is also different. This is also the model used by cv::fisheye in opencv.

Omni-radtan: suitable for most fisheye lenses

Ds/eucm: also used for fisheye lenses with fewer parameters and more compact. But there is also no tangential distortion; the advantage of these two models is that there is a good balance between accuracy and speed, which is more suitable for real-time, optimization-based slam.

For fisheye cameras with large angles, such as 180,210, kalibr may have initialization errors even if the model is appropriate, so you can collect a few more ros packets to try.

If you are in the stage of model selection, you can start with the KB model, that is, you can calibrate it with the pinhole-equi model to check the effect of de-distortion.

At the same time, it should be noted that although KB model and pinhole-fov model are theoretically applicable to fisheye cameras greater than 180 degrees, the tests in kalibr are based on pinhole models, and the results show that these two models are not good for cameras greater than 180 degrees.

Angle > 180 can actually be calibrated using eucm-none.

two。 Notes when calibrating:

When collecting data sets, try to make sure that the calibration board appears in every angle and every corner of the camera image, otherwise the reliability of the calibration result is not high, and initialization errors may occur for cameras with large distortion.

The frame rate recommended by kalibr is 4HZ. Here is a command to lower the frame:

Rosrun topic_tools throttle messages old_topic 4.0 new_topic

With this command, you can reduce the old_topic in the original rosbag to the new_topic of 4hz, and then rosbag play while rosbag record new_topic

At this point, the study of "what to pay attention to when using kalibr calibration" is over. I hope to be able to solve your doubts. The collocation of theory and practice can better help you learn, go and try it! If you want to continue to learn more related knowledge, please continue to follow the website, the editor will continue to work hard to bring you more practical articles!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.