GitPlanet

Cheap and reliable Node.js hosting starts at $3/month, and $1/month static HTML hosting

Created with love in Canada, visit hostnodejs.com today

Feel like to post an Ad? Learn Details
All Projects → google → Draco

google / Draco

Licence: apache-2.0
Draco is a library for compressing and decompressing 3D geometric meshes and point clouds. It is intended to improve the storage and transmission of 3D graphics.

Programming Languages

C++
36643 projects - #6 most used programming language
CMake
9771 projects
javascript
184084 projects - #8 most used programming language
C#
18002 projects
HTML
75241 projects
python
139335 projects - #7 most used programming language

Labels

compressionpoint-cloud3d-graphicsmesh

Projects that are alternatives of or similar to Draco

Polylidar
Polylidar3D - Fast polygon extraction from 3D Data
Stars: ✭ 106 (-97.7%)
Mutual labels:  point-cloud, mesh
Point2Mesh
Meshing Point Clouds with Predicted Intrinsic-Extrinsic Ratio Guidance (ECCV2020)
Stars: ✭ 61 (-98.68%)
Mutual labels:  point-cloud, mesh
Openmvs
open Multi-View Stereo reconstruction library
Stars: ✭ 1,842 (-60.05%)
Mutual labels:  point-cloud, mesh
3d Machine Learning
A resource repository for 3D machine learning
Stars: ✭ 7,405 (+60.59%)
Mutual labels:  point-cloud, mesh
volumentations
Augmentation package for 3d data based on albumentaitons
Stars: ✭ 26 (-99.44%)
Mutual labels:  point-cloud, mesh
Kinectfusionapp
Sample implementation of an application using KinectFusionLib
Stars: ✭ 69 (-98.5%)
Mutual labels:  point-cloud, mesh
vtkplotlib
Wrap up VTK (python 3D graphics library) into an easy to use 3D equivalent of matplotlib
Stars: ✭ 31 (-99.33%)
Mutual labels:  mesh, 3d-graphics
Meshlab
The open source mesh processing system
Stars: ✭ 2,619 (-43.2%)
Mutual labels:  point-cloud, mesh
lepcc
Point Cloud Compression used in i3s Scene Layer Format
Stars: ✭ 22 (-99.52%)
Mutual labels:  compression, point-cloud
pcc geo cnn
Learning Convolutional Transforms for Point Cloud Geometry Compression
Stars: ✭ 44 (-99.05%)
Mutual labels:  compression, point-cloud
Assetkit
🎨 Modern 2D/3D - Importer • Exporter • Util - Library, also called (AssetIO)
Stars: ✭ 97 (-97.9%)
Mutual labels:  3d-graphics, mesh
Point Cloud Utils
A Python library for common tasks on 3D point clouds
Stars: ✭ 281 (-93.91%)
Mutual labels:  point-cloud, mesh
Vedo
A python module for scientific analysis of 3D objects based on VTK and numpy
Stars: ✭ 741 (-83.93%)
Mutual labels:  3d-graphics, mesh
3dhop
3D Heritage Online Presenter
Stars: ✭ 89 (-98.07%)
Mutual labels:  point-cloud, mesh
pcc geo cnn v2
Improved Deep Point Cloud Geometry Compression
Stars: ✭ 55 (-98.81%)
Mutual labels:  compression, point-cloud
Kinectfusionlib
Implementation of the KinectFusion approach in modern C++14 and CUDA
Stars: ✭ 261 (-94.34%)
Mutual labels:  point-cloud, mesh
Easy3d
A lightweight, easy-to-use, and efficient C++ library for processing and rendering 3D data
Stars: ✭ 383 (-91.69%)
Mutual labels:  point-cloud, mesh
Tinyraytracer
A brief computer graphics / rendering course
Stars: ✭ 3,971 (-13.88%)
Mutual labels:  3d-graphics
Recent Stars 2021
🔥🔥🔥SLAM, Pose/Object tracking, Depth/Disparity/Flow Estimation, 3D-graphic, etc. related papers and code
Stars: ✭ 393 (-91.48%)
Mutual labels:  3d-graphics
Ergo
a Framework for creating mesh networks using technologies and design patterns of Erlang/OTP in Golang
Stars: ✭ 376 (-91.85%)
Mutual labels:  mesh
View All Similar Projects ➔

Build Status

News

Attention GStatic users: the Draco team strongly recommends using the versioned URLs for aceessing Draco GStatic content. If you are using the URLs that include the v1/decoders substring within the URL, edge caching and GStatic propagation delays can result in transient errors that can be difficult to diagnose when new Draco releases are launched. To avoid the issue pin your sites to a versioned release.

Version 1.5.0 release

  • Adds the draco_transcoder tool. See the section below on the glTF transcoding tool, and BUILDING.md for build and dependency information.
  • Some changes to configuration variables have been made for this release:
    • The DRACO_GLTF flag has been renamed to DRACO_GLTF_BITSTREAM to help increase understanding of its purpose, which is to limit Draco features to those included in the Draco glTF specification.
    • Variables exported in CMake via draco-config.cmake and find-draco.cmake (formerly FindDraco.cmake) have been renamed. It's unlikely that this impacts any existing projects as the aforementioned files were not formed correctly. See PR775 for full details of the changes.
  • A CMake version file has been added.
  • The CMake install target now uses absolute paths direct from CMake instead of building them using CMAKE_INSTALL_PREFIX. This was done to make Draco easier to use for downstream packagers and should have little to no impact on users picking up Draco from source.
  • Certain MSVC warnings have had their levels changed via compiler flag to reduce the amount of noise output by the MSVC compilers. Set MSVC warning level to 4, or define DRACO_DEBUG_MSVC_WARNINGS at CMake configuration time to restore previous behavior.
  • Bug fixes.

Version 1.4.3 release

Version 1.4.1 release

Version 1.4.0 release

  • WASM and JavaScript decoders are hosted from a static URL.
  • Changed npm modules to use WASM, which increased performance by ~200%.
  • Updated Emscripten to 2.0.
    • This causes the Draco codec modules to return a promise instead of the module directly.
    • Please see the example code on how to handle the promise.
  • Changed NORMAL quantization default to 8.
  • Added new array API to decoder and deprecated DecoderBuffer.
  • Changed WASM/JavaScript behavior of catching exceptions.
  • Code cleanup.
  • Emscripten builds now disable NODEJS_CATCH_EXIT and NODEJS_CATCH_REJECTION.
    • Authors of a CLI tool might want to add their own error handlers.
  • Added Maya plugin builds.
  • Unity plugin builds updated.
  • Bug fixes.

Version 1.3.6 release

  • WASM and JavaScript decoders are now hosted from a static URL
  • Changed web examples to pull Draco decoders from static URL
  • Added new API to Draco WASM decoder, which increased performance by ~15%
  • Decreased Draco WASM decoder size by ~20%
  • Added support for generic and multiple attributes to Draco Unity plug-ins
  • Added new API to Draco Unity, which increased decoder performance by ~15%
  • Changed quantization defaults:
    • POSITION: 11
    • NORMAL: 7
    • TEX_COORD: 10
    • COLOR: 8
    • GENERIC: 8
  • Code cleanup
  • Bug fixes

Version 1.3.5 release

  • Added option to build Draco for Universal Scene Description
  • Code cleanup
  • Bug fixes

Version 1.3.4 release

  • Released Draco Animation code
  • Fixes for Unity
  • Various file location and name changes

Version 1.3.3 release

  • Added ExpertEncoder to the Javascript API
    • Allows developers to set quantization options per attribute id
  • Bug fixes

Version 1.3.2 release

  • Bug fixes

Version 1.3.1 release

  • Fix issue with multiple attributes when skipping an attribute transform

Version 1.3.0 release

  • Improved kD-tree based point cloud encoding
    • Now applicable to point clouds with any number of attributes
    • Support for all integer attribute types and quantized floating point types
  • Improved mesh compression up to 10% (on average ~2%)
    • For meshes, the 1.3.0 bitstream is fully compatible with 1.2.x decoders
  • Improved Javascript API
    • Added support for all signed and unsigned integer types
    • Added support for point clouds to our Javascript encoder API
  • Added support for integer properties to the PLY decoder
  • Bug fixes

Previous releases

https://github.com/google/draco/releases

Description

Draco is a library for compressing and decompressing 3D geometric meshes and point clouds. It is intended to improve the storage and transmission of 3D graphics.

Draco was designed and built for compression efficiency and speed. The code supports compressing points, connectivity information, texture coordinates, color information, normals, and any other generic attributes associated with geometry. With Draco, applications using 3D graphics can be significantly smaller without compromising visual fidelity. For users, this means apps can now be downloaded faster, 3D graphics in the browser can load quicker, and VR and AR scenes can now be transmitted with a fraction of the bandwidth and rendered quickly.

Draco is released as C++ source code that can be used to compress 3D graphics as well as C++ and Javascript decoders for the encoded data.

Contents

Building

See BUILDING for building instructions.

Usage

Unity

For the best information about using Unity with Draco please visit https://github.com/atteneder/DracoUnity

For a simple example of using Unity with Draco see README in the unity folder.

WASM and JavaScript Decoders

It is recommended to always pull your Draco WASM and JavaScript decoders from:

https://www.gstatic.com/draco/v1/decoders/

Users will benefit from having the Draco decoder in cache as more sites start using the static URL.

Command Line Applications

The default target created from the build files will be the draco_encoder and draco_decoder command line applications. Additionally, draco_transcoder is generated when CMake is run with the DRACO_TRANSCODER_SUPPORTED variable set to ON (see BUILDING for more details). For all applications, if you run them without any arguments or -h, the applications will output usage and options.

Encoding Tool

draco_encoder will read OBJ or PLY files as input, and output Draco-encoded files. We have included Stanford's Bunny mesh for testing. The basic command line looks like this:

./draco_encoder -i testdata/bun_zipper.ply -o out.drc

A value of 0 for the quantization parameter will not perform any quantization on the specified attribute. Any value other than 0 will quantize the input values for the specified attribute to that number of bits. For example:

./draco_encoder -i testdata/bun_zipper.ply -o out.drc -qp 14

will quantize the positions to 14 bits (default is 11 for the position coordinates).

In general, the more you quantize your attributes the better compression rate you will get. It is up to your project to decide how much deviation it will tolerate. In general, most projects can set quantization values of about 11 without any noticeable difference in quality.

The compression level (-cl) parameter turns on/off different compression features.

./draco_encoder -i testdata/bun_zipper.ply -o out.drc -cl 8

In general, the highest setting, 10, will have the most compression but worst decompression speed. 0 will have the least compression, but best decompression speed. The default setting is 7.

Encoding Point Clouds

You can encode point cloud data with draco_encoder by specifying the -point_cloud parameter. If you specify the -point_cloud parameter with a mesh input file, draco_encoder will ignore the connectivity data and encode the positions from the mesh file.

./draco_encoder -point_cloud -i testdata/bun_zipper.ply -o out.drc

This command line will encode the mesh input as a point cloud, even though the input might not produce compression that is representative of other point clouds. Specifically, one can expect much better compression rates for larger and denser point clouds.

Decoding Tool

draco_decoder will read Draco files as input, and output OBJ or PLY files. The basic command line looks like this:

./draco_decoder -i in.drc -o out.obj

glTF Transcoding Tool

draco_transcoder can be used to add Draco compression to glTF assets. The basic command line looks like this:

./draco_transcoder -i in.glb -o out.glb

This command line will add geometry compression to all meshes in the in.glb file. Quantization values for different glTF attributes can be specified similarly to the draco_encoder tool. For example -qp can be used to define quantization of the position attribute:

./draco_transcoder -i in.glb -o out.glb -qp 12

C++ Decoder API

If you'd like to add decoding to your applications you will need to include the draco_dec library. In order to use the Draco decoder you need to initialize a DecoderBuffer with the compressed data. Then call DecodeMeshFromBuffer() to return a decoded mesh object or call DecodePointCloudFromBuffer() to return a decoded PointCloud object. For example:

draco::DecoderBuffer buffer;
buffer.Init(data.data(), data.size());

const draco::EncodedGeometryType geom_type =
    draco::GetEncodedGeometryType(&buffer);
if (geom_type == draco::TRIANGULAR_MESH) {
  unique_ptr<draco::Mesh> mesh = draco::DecodeMeshFromBuffer(&buffer);
} else if (geom_type == draco::POINT_CLOUD) {
  unique_ptr<draco::PointCloud> pc = draco::DecodePointCloudFromBuffer(&buffer);
}

Please see src/draco/mesh/mesh.h for the full Mesh class interface and src/draco/point_cloud/point_cloud.h for the full PointCloud class interface.

Javascript Encoder API

The Javascript encoder is located in javascript/draco_encoder.js. The encoder API can be used to compress mesh and point cloud. In order to use the encoder, you need to first create an instance of DracoEncoderModule. Then use this instance to create MeshBuilder and Encoder objects. MeshBuilder is used to construct a mesh from geometry data that could be later compressed by Encoder. First create a mesh object using new encoderModule.Mesh() . Then, use AddFacesToMesh() to add indices to the mesh and use AddFloatAttributeToMesh() to add attribute data to the mesh, e.g. position, normal, color and texture coordinates. After a mesh is constructed, you could then use EncodeMeshToDracoBuffer() to compress the mesh. For example:

const mesh = {
  indices : new Uint32Array(indices),
  vertices : new Float32Array(vertices),
  normals : new Float32Array(normals)
};

const encoderModule = DracoEncoderModule();
const encoder = new encoderModule.Encoder();
const meshBuilder = new encoderModule.MeshBuilder();
const dracoMesh = new encoderModule.Mesh();

const numFaces = mesh.indices.length / 3;
const numPoints = mesh.vertices.length;
meshBuilder.AddFacesToMesh(dracoMesh, numFaces, mesh.indices);

meshBuilder.AddFloatAttributeToMesh(dracoMesh, encoderModule.POSITION,
  numPoints, 3, mesh.vertices);
if (mesh.hasOwnProperty('normals')) {
  meshBuilder.AddFloatAttributeToMesh(
    dracoMesh, encoderModule.NORMAL, numPoints, 3, mesh.normals);
}
if (mesh.hasOwnProperty('colors')) {
  meshBuilder.AddFloatAttributeToMesh(
    dracoMesh, encoderModule.COLOR, numPoints, 3, mesh.colors);
}
if (mesh.hasOwnProperty('texcoords')) {
  meshBuilder.AddFloatAttributeToMesh(
    dracoMesh, encoderModule.TEX_COORD, numPoints, 3, mesh.texcoords);
}

if (method === "edgebreaker") {
  encoder.SetEncodingMethod(encoderModule.MESH_EDGEBREAKER_ENCODING);
} else if (method === "sequential") {
  encoder.SetEncodingMethod(encoderModule.MESH_SEQUENTIAL_ENCODING);
}

const encodedData = new encoderModule.DracoInt8Array();
// Use default encoding setting.
const encodedLen = encoder.EncodeMeshToDracoBuffer(dracoMesh,
                                                   encodedData);
encoderModule.destroy(dracoMesh);
encoderModule.destroy(encoder);
encoderModule.destroy(meshBuilder);

Please see src/draco/javascript/emscripten/draco_web_encoder.idl for the full API.

Javascript Decoder API

The Javascript decoder is located in javascript/draco_decoder.js. The Javascript decoder can decode mesh and point cloud. In order to use the decoder, you must first create an instance of DracoDecoderModule. The instance is then used to create DecoderBuffer and Decoder objects. Set the encoded data in the DecoderBuffer. Then call GetEncodedGeometryType() to identify the type of geometry, e.g. mesh or point cloud. Then call either DecodeBufferToMesh() or DecodeBufferToPointCloud(), which will return a Mesh object or a point cloud. For example:

// Create the Draco decoder.
const decoderModule = DracoDecoderModule();
const buffer = new decoderModule.DecoderBuffer();
buffer.Init(byteArray, byteArray.length);

// Create a buffer to hold the encoded data.
const decoder = new decoderModule.Decoder();
const geometryType = decoder.GetEncodedGeometryType(buffer);

// Decode the encoded geometry.
let outputGeometry;
let status;
if (geometryType == decoderModule.TRIANGULAR_MESH) {
  outputGeometry = new decoderModule.Mesh();
  status = decoder.DecodeBufferToMesh(buffer, outputGeometry);
} else {
  outputGeometry = new decoderModule.PointCloud();
  status = decoder.DecodeBufferToPointCloud(buffer, outputGeometry);
}

// You must explicitly delete objects created from the DracoDecoderModule
// or Decoder.
decoderModule.destroy(outputGeometry);
decoderModule.destroy(decoder);
decoderModule.destroy(buffer);

Please see src/draco/javascript/emscripten/draco_web_decoder.idl for the full API.

Javascript Decoder Performance

The Javascript decoder is built with dynamic memory. This will let the decoder work with all of the compressed data. But this option is not the fastest. Pre-allocating the memory sees about a 2x decoder speed improvement. If you know all of your project's memory requirements, you can turn on static memory by changing CMakeLists.txt accordingly.

Metadata API

Starting from v1.0, Draco provides metadata functionality for encoding data other than geometry. It could be used to encode any custom data along with the geometry. For example, we can enable metadata functionality to encode the name of attributes, name of sub-objects and customized information. For one mesh and point cloud, it can have one top-level geometry metadata class. The top-level metadata then can have hierarchical metadata. Other than that, the top-level metadata can have metadata for each attribute which is called attribute metadata. The attribute metadata should be initialized with the correspondent attribute id within the mesh. The metadata API is provided both in C++ and Javascript. For example, to add metadata in C++:

draco::PointCloud pc;
// Add metadata for the geometry.
std::unique_ptr<draco::GeometryMetadata> metadata =
  std::unique_ptr<draco::GeometryMetadata>(new draco::GeometryMetadata());
metadata->AddEntryString("description", "This is an example.");
pc.AddMetadata(std::move(metadata));

// Add metadata for attributes.
draco::GeometryAttribute pos_att;
pos_att.Init(draco::GeometryAttribute::POSITION, nullptr, 3,
             draco::DT_FLOAT32, false, 12, 0);
const uint32_t pos_att_id = pc.AddAttribute(pos_att, false, 0);

std::unique_ptr<draco::AttributeMetadata> pos_metadata =
    std::unique_ptr<draco::AttributeMetadata>(
        new draco::AttributeMetadata(pos_att_id));
pos_metadata->AddEntryString("name", "position");

// Directly add attribute metadata to geometry.
// You can do this without explicitly add |GeometryMetadata| to mesh.
pc.AddAttributeMetadata(pos_att_id, std::move(pos_metadata));

To read metadata from a geometry in C++:

// Get metadata for the geometry.
const draco::GeometryMetadata *pc_metadata = pc.GetMetadata();

// Request metadata for a specific attribute.
const draco::AttributeMetadata *requested_pos_metadata =
  pc.GetAttributeMetadataByStringEntry("name", "position");

Please see src/draco/metadata and src/draco/point_cloud for the full API.

NPM Package

Draco NPM NodeJS package is located in javascript/npm/draco3d. Please see the doc in the folder for detailed usage.

three.js Renderer Example

Here's an example of a geometric compressed with Draco loaded via a Javascript decoder using the three.js renderer.

Please see the javascript/example/README.md file for more information.

GStatic Javascript Builds

Prebuilt versions of the Emscripten-built Draco javascript decoders are hosted on www.gstatic.com in version labeled directories:

https://www.gstatic.com/draco/versioned/decoders/VERSION/*

As of the v1.4.3 release the files available are:

Support

For questions/comments please email [email protected]

If you have found an error in this library, please file an issue at https://github.com/google/draco/issues

Patches are encouraged, and may be submitted by forking this project and submitting a pull request through GitHub. See CONTRIBUTING for more detail.

License

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

References

Bunny model from Stanford's graphic department https://graphics.stanford.edu/data/3Dscanrep/

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].