Regarding the issue of modifying the input quantity。

linClubs / BEVFusion-ROS-TensorRT

BEVFusion-ROS-TensorRT-CPP real time inference including ros1 & ros2.

BSD 3-Clause "New" or "Revised" License

113 stars 23 forks source link

Closed zyg234 closed 6 months ago

zyg234 commented 8 months ago

Q1：Bevfusion_ At the end of ros.cpp: cv:: Mat img=cv:: imread (pkgpath+"/configurations/cuda bevfusion. jpg"). c Str()

What is the purpose of loading this image。

Q2：What changes should I make if I want to input only one camera (front camera)。

Looking forward to your reply very much！

linClubs commented 8 months ago

Due to the author's limited abilities and the need to quickly publish images with ros， just do it.

You can improve this processing method if it's convenient for you, thanks. ^_^

linClubs commented 8 months ago

It is recommended not to use BEV method with single image, but yolo-xxx.

zyg234 commented 8 months ago

Thank you very much for your reply and suggestions. If I have any improvements, I will let you know.