Skip to content

Rendering LineMod Dataset and Models #1960

Open
@kalperenctnr

Description

@kalperenctnr

I want to render LineMod dataset models and datapoints using pytorch3d but I wasn't able to succeed so far can you help me to point out what part I'm doing wrong?

Thanks.

`import torch
import numpy as np
import matplotlib.pyplot as plt
from pytorch3d.io import load_ply
from pytorch3d.structures import Meshes
from pytorch3d.renderer import (
MeshRenderer, MeshRasterizer, SoftPhongShader, RasterizationSettings,
PointLights, PerspectiveCameras, TexturesVertex
)

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

ply_path = "/home/alperenc/Projects/PoseDatasets/Datasets/lm/lm/models/obj_000001.ply"
verts, faces = load_ply(ply_path)

verts = torch.tensor(verts, dtype=torch.float32, device=device)
faces = torch.tensor(faces, dtype=torch.int64, device=device)

fixed_color = torch.tensor([[0.0, 0.0, 1.0]], dtype=torch.float32, device=device) # Blue color
textures = TexturesVertex(verts_features=fixed_color.expand(verts.shape[0], -1).unsqueeze(0))

mesh = Meshes(verts=[verts], faces=[faces], textures=textures)

R = torch.eye(3, device=device).unsqueeze(0)
T = torch.zeros(1, 3, device=device)

cameras = PerspectiveCameras(device=device, R=R, T=T)

lights = PointLights(device=device, location=[[0.0, 0.0, 3.0]])

raster_settings = RasterizationSettings(image_size=640, blur_radius=0.0, faces_per_pixel=1)

renderer = MeshRenderer(
rasterizer=MeshRasterizer(cameras=cameras, raster_settings=raster_settings),
shader=SoftPhongShader(device=device, cameras=cameras, lights=lights)
)

image = renderer(mesh)
image = image[0, ..., :3].cpu().numpy()

plt.imshow(image)
plt.axis("off")
plt.show()
`

Output is all blue image

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions