Generate a 3D BoundingBox from 2D Segmentation · Issue #4929 · isl-org/Open3D · GitHub
I've been using Open3D for some of my 3D work and I love the software. But I am still fairly new to it. I don't know if this question is simple but I would appreciate any guide and help on implementing it. I am trying to generate a 3D BoundingBox in a PointCloud with respect to a 2D Segmentation mask of my original RGB image (PointCloud being created from an RGB-D image). For example, I have the following RGB-D image and a segmented object mask: These are the steps of what I'm thinking of doing: 1. Create an RGBDImage object with the create_from_color_and_depth function. 2. Create a PointCloud object with RGBDImage with the create_from_rgbd_image function. 3. Create a Vector3dVector object from the list of points with respect to the 2D object mask. --> This is mainly the part I don't know how to do. 4. Generate an AxisAlignedBoundingBox from the Vector3dVector object using the create_from_points function. 5. Visualize the BoundingBox and my PointCloud using the visualization.draw. Thes
Explore this link on the map →saved by
related reading
- Convex hull, Image processing, Image Classification, Image retrieval, Shape detectiondiva-portal.org
- 3D-LFM: Lifting Foundation Model3dlfm.github.io
- [2403.14530] HAC: Hash-grid Assisted Context for 3D Gaussian Splatting Compressionarxiv.org
- Ray Tracing in One Weekendraytracing.github.io
- ReferIt3D: Neural Listeners for Fine-Grained 3D Object Identification in Real-World Scenesecva.net
- Claude Fable 5 Creates Minecraft Clones & 3D Worlds from Prompts | explainx.ai Blog | explainx.aiexplainx.ai
- [1711.06396] VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detectionarxiv.org
- 3DMV3dmv2023.github.io
- Language-Grounded Indoor 3D Semantic Segmentation in the Wildarxiv.org
- segment-geospatialsamgeo.gishub.org
- PyVista | 3D plotting & analysis made easypyvista.org
- LearnOpenGL - Hello Trianglelearnopengl.com