Virtual camera problem

[onionysy]

This method uses virtual camera, which is a good idea. However, in the data processing stage, only the image is mapped to the virtual camera, and no corresponding transformation is done to the three-dimensional coordinate points.

[onionysy]

In my understanding, when using the virtual camera, the coordinates are converted to the virtual camera coordinate system, not the original camera coordinate system. However, the 3D coordinates['xyz'] read in 'Dataloader' are still in the original camera coordinate system and should be converted to the virtual camera coordinate system. In fact, there is no such process during Dataloader. What is this because of?

[qinjian623]

Hi,

The points in the 3D space are unified in the vehicle coordinate system, so there is no need to perform spatial mapping again. The VC is responsible for unifying the image pixel space.

[EnternalTwinkle]

@onionysy Can you share the source code? Thank you very much，1017094591@qq.com

[FeiyuLi1]

@onionysy Can you share the source code? Thank you very much，2235104161@qq.com