ROS2:立体声图像纠正

发布于 2025-01-28 08:55:00 字数 767 浏览 3 评论 0原文

我开始开发一个视觉探射程序。 “前”部分(图像采集和希望纠正)是在ROS2中完成的,然后使用OpenCV完成处理的核心。

我现在从一个ROS袋(直到获得真实的相机)以及相应的校准信息中获取立体声图像。我有以下主题:

/stereo/left/image_raw
/stereo/left/camera_info
/stereo/right/image_raw
/stereo/right/camera_info

我设法在自己的rosnode中检索了图像和camera_info(即使我还没有实现该部分,我也应该能够将它们转换为OpenCV图像)。

但是,我的图像管道中有2个丢失的步骤:不隔离和立体调整。

在ROS1中,我本可以从“ stereo_image_proc”软件包中使用“ stereo_image_proc”节点来执行这两个步骤:

ROS_NAMESPACE=stereo rosrun stereo_image_proc stereo_image_proc

但是,在ROS2中,事实证明,stereo_image_proc仍然存在,但是stereo_image_image_proc节点不再存在。

您是否知道任何ROS2软件包,可以将一对RAW CAMARA图像主题(+ Camera_info主题)转换为未置换和整流的图像? (颜色或BW的结果,我不太在意)

或者如果更容易,是否有一种(简单)在OpenCV中执行这种不合理/整流的方法(使用Camera_info主题中的数据)?

非常感谢

I'm starting to develop a visual odometry program.
The "front" part (image acquisition and hopefully rectification) is done in ROS2, then the core of the processing will be done with openCV.

I'm getting the stereo images for now from a ros bag (until I get the real camera), along with the corresponding calibration information. I have the following topics :

/stereo/left/image_raw
/stereo/left/camera_info
/stereo/right/image_raw
/stereo/right/camera_info

I managed to retrieve the images and camera_info in my own rosnode (and I should be able to transform them to opencv images, even if I haven't implemented that part yet).

However, I have 2 missing steps in my image pipeline : undistorting and stereo-rectification.

In ROS1, I could have used the "stereo_image_proc" node from the "stereo_image_proc" package to do those 2 steps :

ROS_NAMESPACE=stereo rosrun stereo_image_proc stereo_image_proc

In ROS2 however, it turns out that the package stereo_image_proc still exists, however the stereo_image_proc node in it no longer exists.

Do you know about any ROS2 package that enables transforming the pair of raw camara image topics (+ camera_info topics) into undistorted and rectified images? (results in color or BW, I don't care much)

Or if it is easier, is there a (simple) way to do this undistortion/rectification in opencv (using the data from the camera_info topics)?

Thanks a lot in advance

如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。

扫码二维码加入Web技术交流群

发布评论

需要 登录 才能够评论, 你可以免费 注册 一个本站的账号。

评论(1

半透明的墙 2025-02-04 08:55:00

In the comments of this tutorial someone suggested using image_proc for rectified image creation 。
我尝试了它,图像看起来很纠正,但是在我的情况下,问题是 /差异主题是空的。
这是图形 https://paste.pics.pics/jy08e
I also asked a question is ROS答案

In the comments of this tutorial someone suggested using image_proc for rectified image creation.
I tried it and the images looked rectified, but the problem in my case is that the /disparity topic is empty.
Here is the graph https://paste.pics/JY08E
I also asked a question is ROS Answers

~没有更多了~
我们使用 Cookies 和其他技术来定制您的体验包括您的登录状态等。通过阅读我们的 隐私政策 了解更多相关信息。 单击 接受 或继续使用网站,即表示您同意使用 Cookies 和您的相关数据。
原文