shichaoy / semantic_3d_mapping Goto Github PK

First, thank you for sharing the code?
Could you tell me how to get the "CameraTrajectory.txt"? and meaning of parameters of each row.
And I want to know how to compute the camera pos？
Thanks!

In compute() function of the ScrollForBaseFrame class, there exists the following code:

tf::Vector3 target_sensor_to_center = wv2laser * ca::point_cast< tf::Vector3 >(target_sensor_to_center_); Vec3 origin_laser = ca::point_cast<Vec3>(target_sensor_to_center); //sensor position in world frame.
What is the meaning of "target_sensor_to_center_" , "wv2laser" and "origin_laser" ?
and what is the meaning of "origin_laser-center"?

Hi, Thanks for the interest. I think it is due to recent server changes. I put it temporarily here. Refer to sh file to put it in right place. https://drive.google.com/open?id=1cfQYQMkU1cZn6e4hPsgGyYwvKfh7gYV7

I will update the sh file when I find a stable position.

Originally posted by @shichaoy in #6 (comment)

unzip: cannot find or open data_kitti.zip, data_kitti.zip.zip or data_kitti.zip.ZIP.

Hello,
I want to test your research on a real drone, and I'm wondering that images from monocular camera can be used on your research, because semantic segmentation can be built from RGB images, and some light weight CNNs can run in real time on embbeded system (like you wrote in your paper). Can you answer this for me?
Thank you and best regard.

I have run your script with matlab but it has error ?

change_cnn_label
processing 1 out of 500
Error using reshape
To RESHAPE the number of elements must not change.

Error in change_cnn_label (line 38)
prob=reshape(U,num_labels,img_height*img_width)';
Can you give me some advice to fix this ?

I noticed in the code grid_sensor/src/main_stereo_imgs.cpp, it writes the reprojected images for evaluation. But where can we find the ground truth? The website of Kundu14 seems to be down.
Thnaks!

Could you please explain a little bit how did you gain the pose estimate using ORB SLAM2? Thanks!

Hi, when we trying to run kitti15, our 3D scrolling map seemed to be drifting. We are not sure if we generate the depth images and the orb slam poses. Currently, we generate them using kitti's gray camera (and its calibration matrix). When you generate kitti05's depth images and poses, did you convert the rgb left camera into gray, or you directly use the gray camera?

Thanks!

Hi, thanks for share!
Could you please tell me how to save the final semantic map like this?

Thanks!!

Hi, thanks for share!
I was confused that what the data struct in ./grid_sensor/data_kitti/superpixel_bin/000000.bin and how to create these files?
Thanks!

(1) we download the kitti gray sequences, because they are stereo, we firstly use the elas to convert stereo image into depth. one question, I saw the depth data you provided have the following property: close objects are dark, far objects are light. when I use the elas, how I can save the depth as the format you have?

Thank you for sharing your code.

I can not download data_kitti.zip at http://www.frc.ri.cmu.edu/~syang/dataset/kitti_crf/data_kitti.zip
Could you give another link to download this data_kitti.zip?

Hi, thanks for your share!

I am testing your code on the NYUv2 dataset. ^ _ ^

I have used your code in folder named 'preproess_data/superpixel/' and have made the image like this

Then transform the image to bin using your code.

I make the semantic segmentation using the FCN, in which there is model on the dataset.
The segmentation image like this

Then translate the image into bin using the code:
net.blobs['prob'].data[0].transpose([1,2,0]).flatten().tofile('/home/x/bin/nyu'+ str(i) + '.bin')

And I get the pose through the ORB-SLAM2

Then I run the mapping_3d, there isn't any warning and error.

However, the result seems to be wrong. -_-
The image produced in the folder named 'crf_3d_reproj' seems to be wrong, like this

As a result, the semantic 3d mapping looks bad and like this

Could you please help me and tell me what the possible reasons are ?

Thanks!

Hello! I have realized that there are only 20 depth images in your data set. And it complains about can not read depth image when the index goes to 20

When calling predict_dilanet.py , complained as follow:

To processing: 4
First one: 000001.png
Traceback (most recent call last):
File "predict_dilanet.py", line 223, in
batch_predict_kitti("kitti")
File "predict_dilanet.py", line 126, in batch_predict_kitti
dataset = Dataset(dataset_name)
File "predict_dilanet.py", line 30, in init
.format(dataset_name))
IOError: Do not have information for dataset kitti

could you please explain a little bit what this may would have caused?

Thanks!

HI I wonder if you have any idea where the things could have gone wrong, as you can see that things here are not parallel....

Hi, thanks for share!
how can i get the evaluatioList.txt
i can run without the evaluatioList.txt
what is the function of it
@shichaoy