- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
Hi, I'm trying to save a PC2 and convert it to Image, but it says there's no RGB info. Does the Euclid only provide XYZ info or am I missing something?
If there's actually no RGB fields, how can I convert it on a grayscale?
Thanks!
コピーされたリンク
- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
Hi Georgge,
Could you please be more specific? What is PC2 exactly and how are you trying to convert it to image?
Regards,
Diego V.
- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
Hi Diego,
First of all, thanks for answering. PC2 was an abbreviation for PointCloud2, and all I'm really trying to do is the following:
-Run /camera/depth/points
-Save one frame in a .pcd with "rosrun pcl_ros pointcloud_to_pcd"
-Publish that .pcd in some visualizer to be sure that it did its work properly.
I saw that the /camera/depth/image_transcoded topic is a sensor_msg/Image one, but I assume it gets the info from the two stereo IR sensor, just as /camera/depth/points. Of course the transcoded subscribes to the ../image_raw, so all I'm asking is: is there any relationship between sensor_msg/Image and sensor_msg/PointCloud2? Because it seems that the Euclid is able to publish several topics with the info of the same hardware (in this case IR sensor). How could I transform the PointCloud2 into an Image, knowing that the Euclid has all the information it needs to publish depth topics in different types of messages? Do I need more than a .pcd (which consists of XYZ values) or does it suffice for my purpose of converting it to an image?
- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
Hello Georgge,
I am part of the Engineering team, and i will gladly try to help you, I understand what you are trying to do, but to clarify you want to save a point cloud + RGB scene into an image ?
Euclid already publishes the RGB frame and the PC2 Frame Calibrated. you can visualize the 3D pointcloud +RGB in RVIZ. regarding saving the frame as an Image i will have to take a look at it. and get back to you ASAP.
- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
Poor help from Intel, guys. I sent you an email almost one month ago and still no help.
Should reconsider this. Thanks
- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
Hello Georgge,
Euclid is a development platform meant for users to develop their own applications on top of. Intel gives support on the functionality of Euclid.
Euclid exposes the RGB stream as well as the depth stream. You can write a ROS node to register the corresponding data and save it to whatever you want. Another option would be to use a ROS bag and record the data.
Best regards,
Jesus G.
Intel Customer Support
- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
If you decide to follow the suggestion of Jesus to record the data into a ROS bag file, then you may be able to do so with a ROS module called rqt_bag.
http://wiki.ros.org/rqt_bag rqt_bag - ROS Wiki
The record button is the very first icon on the toolbar, a circle that turns red when clicked on to indicate that recording is taking place. Recording is then stopped by clicking on the circle button again.
- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
Hi Georgge,
You can write a rosnode that subscribes to the topic and saves the messages to a file. I'll be posting a code sample for doing exactly that sometime in the next few weeks that you could use if you still find yourself struggling with it.
That might be the better way to save it for post processing via matlab or the like.
Regards,
Meitav
Intel Euclid Development Team
- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
Hi Georgge,
/camera/depth/image_transcoded gives reformatted data from /camera/depth/image_raw , because it needs to be in a different format to properly display in the browser.
You can subscribe to the depth images' topic and write a program for plane segmentation, but we don't provide code that does that at this time.
Regards,
Meitav
Intel Euclid Development Team
- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
AFAIK, the image_raw topic gives a measurement of the distance in mm of every pixel. I assume I can export all those pixels and know the distance to the Euclid. I've read a bit and I think there are few ways to do it, but the most common one is converting the ROS image to CV image. Is that the best way or is there a fastest one?
Regards,
George
