search

DeyvidKochanov-TomTom / kprnet

MIT License
65 stars 12 forks source link

new_raw data,thanks #20

Open emilyemliyM opened 2 years ago

emilyemliyM commented 2 years ago

Dear author, I want to know about the "0.2" &"0.8", how to choose the value. when I use my own dataset to do the range projection operation, I just got the XY postion between(2049,69) So I really want to know how to set the value to get a new raw!!

thanks!

image

DeyvidKochanov-TomTom commented 2 years ago

Hi, this you can checkout this issue #8 and #9 for more details. The current method is really designed to work for kitti and how you implement the projection will depends on the dataset. If you give me more details I might be able to help.

emilyemliyM commented 2 years ago

您好， 您可以查看此问题#8和#9了解更多详细信息。当前方法实际上是为 kitti 设计的，您如何实现投影将取决于数据集。如果你给我更多的细节，我可能会提供帮助。

Dear author, thanks for the quick reply!!

In order to quickly run through the data flow, the dataset I used is the semantic kitti dataset, and then, I used data augmentation, adding hundreds of point cloud data to each scan in the training phase.

There are no other operations related to data preprocessing. But you will find that after range mapping, some coordinates are greater than 65, some are less than, and some are equal.

thanks！