generating new images from woodscape dataset

[Ichrak-shili]

Hi,
Thank you for providing the code.
I'm trying to generate new images from woodscape with labels for object detection. I create a new calibration file with new distortion parameters, new width and hight. I change the cx_offset and cy_offset by multiplying with new_width/width and new_height/height respectively. I create two cameras instances using create_img_projection_maps method from projection.py.
For the new bounding boxes, I just use 2d_to_3d from src camera and 3d_to_2d from dest camera.
I get some artifacts around the center of the image and the bouding boxes don't seem to be correct.
Do I have to change other things in the calibration file or make other steps please ?

I tested with new coefficients: [32.13785692, 81.25326978, 38.84755383, 2.82090108], new_width=950, new_height=700
The results on '07848_MVL' image are: