mli0603 / blenderneuralangelo Goto Github PK

Hey there, thanks for creating this tool, it's been a really exciting one to start working with.

I've been trying to create a tight bounding box for a head but I'm having problems making it happen. After running the COLMAP but before generating the JSON file for data loading, I loaded up my COLMAP data in blender and the bounding box seemed reasonably well sized (as seen in the photo below).

The moment I try to change it's size though it becomes massive and no matter how much I scale it down it's still way too big for my model.

I've also noticed a few of my point clouds are very unaligned with my camera, I've included those below for reference. Please let me know if this is a glitch or if there may be a way to make this work!

Thanks :)

Whenever you try installing it, it just doesn't, when I run it as a script within the last versions of blender it says the following:

Python: File "\neuralangelo_addon.py", line 95
<title>BlenderNeuralangelo/neuralangelo_addon.py at main · mli0603/BlenderNeuralangelo</title>
^
SyntaxError: invalid character '·' (U+00B7)

I tried erasing that character, then this error pops up:

Python: File "\neuralangelo_addon.py", line 7

^
SyntaxError: invalid syntax

I guess when you removed dense folder, there are no images in the repo anymore. Can you add the toy example images?

Can only generate boundary spheres, and is it possible for boundary cubes?

hi, I wanna know that how long does neualangelo takes for a scene? maybe 12hours?

besides, how did you render the mesh with a given pose, blender?

And I don't know how to run this instrucion:
./blender --python PATH_TO_BlenderNeuralangelo/start_blender_with_addon.py

First, thanks for your work.

I add the BlenderNeuralangelo addon to Blender
I try to load toy_example, when I press the button load COLMAP Data, Python is reporting an error.

• I can see the pointcloud, but there is no image in image plane.

• And I can't switch camera views in the timeline, it seems that camera extrinsics is not loaded sucessfully

• In the folder 'toy_example/dense', there is blender_images, and I can see 17 images in it.

Python: Traceback (most recent call last): File "C:\Users\Administrator\AppData\Roaming\Blender Foundation\Blender\3.6\scripts\addons\neuralangelo_addon.py", line 1113, in execute load_camera(colmap_data, context) File "C:\Users\Administrator\AppData\Roaming\Blender Foundation\Blender\3.6\scripts\addons\neuralangelo_addon.py", line 963, in load_camera generate_camera_plane_texture(image_sequence) File "C:\Users\Administrator\AppData\Roaming\Blender Foundation\Blender\3.6\scripts\addons\neuralangelo_addon.py", line 891, in generate_camera_plane_texture material.node_tree.links.new(image_texture.outputs['Color'], principled_bsdf.inputs['Base Color']) AttributeError: 'NoneType' object has no attribute 'inputs'

when I load COLMAP Data,there is an error
Python: Traceback (most recent call last):File"neuralangelo addon.py", line 1095, in executeload camera(colmap data, context)File "neuralangelo addon.py", line 969 , in load camneragenerate camera plane texturelimage sequence)File "neuralangelo addon.py, line 897 , in generate camera plane texturematerial.node tree.links.new(image texture.outputs[ 'Color' ], principled bsdf inputs[Base Color'])AttributeError: 'NoneType' object has no attribute 'inputs'
the version of blender is 3.5.1

Do you have any plans to make a demo available? Thank you very much for your time

When I try to load the example one the error was:

Python: Traceback (most recent call last):
File "\neuralangelo_addon.py", line 1095, in execute
load_camera(colmap_data, context)
File "\neuralangelo_addon.py", line 950, in load_camera
bpy.ops.image.open(filepath=blender_img_path, directory=blender_img_path, files=blender_file_names_formatted,
File "C:\Program Files\WindowsApps\BlenderFoundation.Blender_3.5.1.0_x64__ppwjx1n5r4v9t\Blender\3.5\scripts\modules\bpy\ops.py", line 113, in call
ret = _op_call(self.idname_py(), None, kw)
TypeError: Converting py args to operator properties: IMAGE_OT_open.files error converting a member of a collection from a dicts into an RNA collection, failed with: TypeError: Converting a Python list to an RNA collection: keyword "type" missing

Thanks for the great work and open-sourcing this useful tool.

It seems that the tool only needs sparse point clouds, camera poses, and images. The image distortion should be negligible for deciding bounding boxes. Therefore, I have the feeling that we do not need dense reconstruction which might take a long time to run.

In that case, is it enough to just specify the image folder and the sparse folder for this addon?