kujason / ip_basic Goto Github PK

The results for bilateral blurring in Table III were entered incorrectly. They should be:

 Kernel                 | RMSE (mm) | MAE (mm) | Runtime (s)
Bilateral Blur          |   1503.67 |   310.59 | 0.011 
Median + Bilateral Blur |   1454.75 |   303.71 | 0.014

Hello, suppose I obtained a sparse depth image by projecting 3D laser onto an image. I'm wondering if it's possible to use your method to convert the sparse depth image into a dense depth image.

Thanks to the author for a job well done. When I tried to use the files under devkit/cpp/ for effect evaluation, I found that the results obtained were incorrect. As shown below, I don’t know what’s going on. Hope to get your help and look forward to your reply

Thanks for your great idea!

Could you tell me what is the noise removal algorithm showed in the youtube video? (The right sample in the second row)
https://www.youtube.com/watch?v=t_CGGUE2kEM&t=3s&ab_channel=WAVELaboratory

Thanks!

Hi, wonderful work! I want to ask, based on your work, the sparsity of lidar can be relieved, and can it imporve the detection performance of lidar? like in avod?

@kujason
Hi, I am trying to using your repo.
I found that the IP_basic accept .png file.
How to using .bin file by ip-basic? (Or, can you help me convert velo into depth-map? thanks)

Thank you for sharing your code!
I would just like to know the possibility of applying your code on normal depth images, like depth map image of kinect or stereo! the idea is to fill the invalid pixels (black pixels in attached image ) .
Thank you in advance !

Now my background color is black. However, the further the object is, the whiter the pixels should be. So, it more makes sense if the background color is white.

On the other hand, how to change the relationship between depth and color shade? It also makes sense if the further the object is, the blacker the pixels are.

Hi,

I'm trying to apply your algorithm to my dataset which is also semi-sparse lidar point image with resolution 1280*760. Do I need to make some adjustment to your code? Since I run the demo code on my image and the result is not satisfying.

Thank you!

I follow the Setup but end up with [ModuleNotFoundError: No module named 'ip_basic'].
How can I solve this problem and run the code?
Waiting for ur reply, thx

Hello,

I want to ask, How did you colorized the pointcloud?
like in this image
https://github.com/kujason/ip_basic/raw/master/images/006338/lidar.png

数据集下载有链接吗？

Hello! As in the topic. Could anyone, please, tell me how this solution performs on lower-beam lidars, especially 16-beam Velodyne?

Hi,

Have you experimented with combining this method with also RGB information? At the moment depth completion is just using the depth itself.

Cheers

I'm already run this demo with zero error and get the output, but I dont know why my output folder is empty, the Kitti folder is organized as yours. waiting for your reply and really thank u.