Open victorhuzhening opened 11 months ago
Hi,
if your model DOES specify a PINHOLE camera, the above error shouldn't be raised, so i strongly suspect that it uses something else.
This is because there are some distortion in your images. Fortunately,Colmap has a function to solve it. Colmap->Reconstruction->Dense reconstruction->Undistortion. After those processes,3DGS can run correctly. I hope this will give some help.
@PKU-YQH I tried you suggestion, and change the images path to the new "images" folder that were created after clicking "Undistortion". GS train.py still does not work. How did you get it to work?
@codegrafix Did you figure this out? I keep getting
(gaussian_splatting) C:\Users\Demon\Nerf Guassian\gaussian-splatting>python convert.py -s data/input [option_manager.cc:815] Check failed: ExistsDir(*image_path) ERROR: Invalid options provided. ERROR:root:Feature extraction failed with code 1. Exiting.
same
This is because there are some distortion in your images. Fortunately,Colmap has a function to solve it. Colmap->Reconstruction->Dense reconstruction->Undistortion. After those processes,3DGS can run correctly. I hope this will give some help.
what did you do after Undistortion
same error here..
@cucdengjunli @RoyAmoyal
@cucdengjunli @RoyAmoyal
this works, thx!
I dont understand the solution given by @cucdengjunli @RoyAmoyal can someone please explain in detail what to do
Sry, by @luojin , can someone please explain what to do exactly @GaussRieman , how did you get it resolved?
I have a question about the camera model of the data: is cameras.bin where the gaussian_splatting model reads the model from? I have a cameras.bin file that I believe specifies PINHOLE model, but I keep getting this error:
Optimizing Output folder: ./output/752b4cb7-a [03/10 04:25:09] Reading camera 1/315Traceback (most recent call last): File "/content/gaussian-splatting/train.py", line 216, in
training(lp.extract(args), op.extract(args), pp.extract(args), args.test_iterations, args.save_iterations, args.checkpoint_iterations, args.start_checkpoint, args.debug_from)
File "/content/gaussian-splatting/train.py", line 35, in training
scene = Scene(dataset, gaussians)
File "/content/gaussian-splatting/scene/init.py", line 44, in init
scene_info = sceneLoadTypeCallbacks["Colmap"](args.source_path, args.images, args.eval)
File "/content/gaussian-splatting/scene/dataset_readers.py", line 145, in readColmapSceneInfo
cam_infos_unsorted = readColmapCameras(cam_extrinsics=cam_extrinsics, cam_intrinsics=cam_intrinsics, images_folder=os.path.join(path, reading_dir))
File "/content/gaussian-splatting/scene/dataset_readers.py", line 95, in readColmapCameras
assert False, "Colmap camera model not handled: only undistorted datasets (PINHOLE or SIMPLE_PINHOLE cameras) supported!"
AssertionError: Colmap camera model not handled: only undistorted datasets (PINHOLE or SIMPLE_PINHOLE cameras) supported!
I also have the files organized as shown in the README. Any clues/tips on what might be causing the problem? Should I maybe specify camera model when I generate the files from Colmap?
Appreciate the help.