3D Point Cloud Visualizer (PLY, ...)

A Visual Studio Code extension for visualizing Point Cloud files like PLY (Polygon File Format) with interactive 3D visualization using Three.js. Multiple 3d formats for points and shapes and depth and disparity images are also supported. The depth and disparity images are converted on the fly into point clouds.

Frequently new features will be added. You can simply request features that support your workflow or new file formats on github.

Features

• Visualize Point Clouds: Navigate in colored or uncolored point clouds
• Fast Loading of Big Point Clouds: Even point clouds with 5 Million Points load in around a second
• Compare Multiple Point Clouds: Multiple point clouds can be loaded in the same view and activated and deactivated independently
• Rotation of points: Apply Rotation Matrix for single point clouds
• Point Cloud Formats: Support for both ASCII and binary PLY point clouds and meshes, XYZ, XYZN, XYZRGB, PCD and PTS point clouds, OBJ, STL, OFF and GLTF/GLB meshes
• Depth to point cloud conversion: Can convert on the fly depth images to point clouds. It has support for tiff, png, pfm, npy and npz files
• Depth to point cloud settings: Allow setting the fx, fy, cx, cy, different camera disturbance models, mono depth scale and bias parameter, png int16 scale parameter, disparity offset ...
• Pose estimations: Currently in development is support for 3d pose estimations in json files
• Buttons for meshes: Easily active wireframe, mesh, points and normals
• Corresponding website: All features can be also used without VS Code using the static website https://f-kleinicke.de
• Frame rate visualization: Shows the current frame rate. When point cloud is not moved, no more frames will be generated to save power, also ignores transparencies by default to save power

Interpret depth/disparity images as point cloud.

Load two point clouds and switch between them imediatly pressing shift and click

Roadmap

• Wrong direction of Rotation: When rotating the point cloud the direction is inverted. It's unclear how to fix this.
• Add eye dome lighting: Improve perception of uncolored point clouds
• Drag and Drop: Add more pc to a given pc using drag and drop instead of adding them in an extra window
• Add new file formats: Body pose json files for 3d body pose, FBX files can contain meshes or animations, Exr files can contain float depth images or complete point clouds, gaussian splats
• Measure distance: Add option to measure the distance between points
• Dataset support: Add images from Middlebury stereo and Eth3d as example images
• Automatic usage of calib files: calibration files next to the depth images can contain all required parameter for a depth to pc conversion. Automatically use those for depth images for intrinsics and extrinsics for point clouds
• Improve npz file interpretation: These files can have different structures. Offer some options to visualize them

Feature Requests and Issues

If you have use cases that would be helpful for others or find problems, feel free to suggest them on the GitHub repository. If you know how to fix bugs or how to implement certain features, feel free to contribute. If you provide an example file, I can implement support for it in the extension.

Appendix:

Required structures of certain file formats

npy

• As depth image: [X,Y]
• As point cloud: [...,3] with the 3 values X,Y,Z