search

llreda / Stereo_Matching

depth estimation on the SCARED dataset.
5 stars 0 forks source link

processing.py and the method of how to get the visualization #1

Open Tobyzai opened 1 year ago

Tobyzai commented 1 year ago

Hi llreda, thanks for sharing such a helpful project.

I have two questions:

  1. when using processing.py to get training data and test data, it takes up a lot of space (for example, the unpacked dataset_2 takes up 365G of storage space), is this actually the case? Is there a solution for this?
  2. What is the method used to generate the visualization in the final GIF motion picture?

I hope to receive your feedback.

llreda commented 1 year ago

It's a pleasure to receive your question，

  1. The decompressed data set is really large, up to about 1TB， I don't do any additional processing here. Because the tiff file is too large, you can try to store it in 16 bit png or other file formats.
  2. I use PCL library to visualize and record the screen, and then convert it into GIF.

---- Replied Message ---- | From | @.> | | Date | 10/16/2022 18:02 | | To | @.> | | Cc | @.***> | | Subject | [llreda/Stereo_Matching] processing.py and the method of how to get the visualization (Issue #1) |

Hi llreda, thanks for sharing such a helpful project.

I have two questions:

when using processing.py to get training data and test data, it takes up a lot of space (for example, the unpacked dataset_2 takes up 365G of storage space), is this actually the case? Is there a solution for this? What is the method used to generate the visualization in the final GIF motion picture?

I hope to receive your feedback.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you are subscribed to this thread.Message ID: @.***>

Tobyzai commented 1 year ago

Thank you so much for such a quick reply. Could you elaborate a bit on how to visualize the SCARED2019 dataset using the PCL library? Is it visualized using color images (.png) and the corresponding point cloud images (.tiff)? Or where can I get more information about it? I'm really sorry, I'm new to this area and still need to keep learning. Thank you for your help.

llreda commented 1 year ago

Yes, you only need to use color images and point cloud information to visualize. Using PCL's PointXYZRGB class, you can refer to https://pcl.readthedocs.io/projects/tutorials/en/latest/pcl_visualizer.html?highlight=pcl_visualizer_demo#compiling -And running the program
I'm sorry that I can't find the code at that time.

At 2022-10-16 20:37:06, "Tobyzai" @.***> wrote:

Thank you so much for such a quick reply. Could you elaborate a bit on how to visualize the SCARED2019 dataset using the PCL library? Is it visualized using color images (.png) and the corresponding point cloud images (.tiff)? Or where can I get more information about it? I'm really sorry, I'm new to this area and still need to keep learning. Thank you for your help.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>

Tobyzai commented 1 year ago

Thank you very much. I will try it. Since I am new to the field of 3D reconstruction related, I would appreciate if you could give me some advice on how to learn? What should I learn?