During the Image Undistortion step of the process, I get the following error when attempting to use the simple pinhole camera model and the same camera parameters used in the demo:
Traceback (most recent call last):
File "main.py", line 13, in <module>
dp.process(params)
File "<path to>/consistent_depth/process.py", line 117, in process
return self.pipeline(params)
File "<path to>/consistent_depth/process.py", line 62, in pipeline
valid_frames = calibrate_scale(self.video, self.out_dir, frame_range, params)
File "<path to>/consistent_depth/scale_calibration.py", line 184, in calibrate_scale
video.path, colmap.sparse_dir(colmap_dir, 0)
File "<path to>/consistent_depth/scale_calibration.py", line 79, in make_camera_params_from_colmap
cameras, images, size_new
File "<path to>/consistent_depth/utils/load_colmap.py", line 175, in convert_calibration
intrinsics = cameras_to_intrinsics(cameras, sorted_cam_ids, size_new)
File "<path to>/consistent_depth/utils/load_colmap.py", line 116, in cameras_to_intrinsics
for c in cameras.values()))
AssertionError
All the line throwing the error is checking is if the model is:
assert all(
(c.model == "SIMPLE_PINHOLE" or c.model == "PINHOLE"
or c.model == "SIMPLE_RADIAL"
for c in cameras.values()))
which it seems like I am complying to? my full command is sudo python3 main.py --video_file <path_to_avi> --path <path_to_results> --camera_model "SIMPLE_PINHOLE" --camera_params "1671.770118, 540, 960" --make_video where the video being used in my command is not the same video as the one in the demo. Am I missing something? I know the camera params are probably off for my use case, but I imagine it should run to completion regardless?
turned out to be my mistake: an earlier checkpoint using an incorrect camera model (in the same output directory) was causing issues with the new run. solution was clearing the output directory
During the Image Undistortion step of the process, I get the following error when attempting to use the simple pinhole camera model and the same camera parameters used in the demo:
All the line throwing the error is checking is if the model is:
which it seems like I am complying to? my full command is
sudo python3 main.py --video_file <path_to_avi> --path <path_to_results> --camera_model "SIMPLE_PINHOLE" --camera_params "1671.770118, 540, 960" --make_video
where the video being used in my command is not the same video as the one in the demo. Am I missing something? I know the camera params are probably off for my use case, but I imagine it should run to completion regardless?