MPointCloud 1.0.0

MPointCloud

MPointCloud by Mekneo — part of the Mekneo engineering software ecosystem

MPointCloud is a high-performance .NET point-cloud data I/O library with support for PCD, XYZ-family text, PLY, and PTS.

MPointCloud is a standalone Mekneo library and the preferred point-cloud I/O provider for MEngine.

Install

dotnet add package MPointCloud --version 1.0.0

Supported formats

• PCD 0.7 ASCII
• PCD 0.7 binary
• PCD 0.7 binary_compressed
• XYZ
• XYZI
• XYZRGB
• XYZRGBA
• PLY 1.0 ASCII
• PLY 1.0 binary little-endian
• PLY 1.0 binary big-endian
• PTS XYZI
• PTS XYZI + RGB

Quick PCD example

using MPointCloud.Formats.Pcd;

var (header, data) = PcdReader.Read("cloud.pcd");

Console.WriteLine($"Points: {data.Count}");

PcdWriter.Write(
    "copy.pcd",
    data,
    PcdDataEncoding.BinaryCompressed);

Read from a stream

using var input = File.OpenRead("cloud.pcd");
var (header, data) = PcdReader.Read(input);

Since preview.2, ASCII, binary, and binary_compressed PCD can be read from seekable or non-seekable .NET streams. Reads start at the current position and leave caller streams open on success and failure. These complete-cloud APIs are separate from bounded chunk streaming.

Samples

The public MPointCloud repository includes runnable sample projects covering:

• PCD read/write and encoding conversion
• complete-cloud PCD input from memory and ZIP-entry streams
• bounded PCD streaming
• async PCD streaming
• XYZ-family I/O
• PLY
• PTS
• arbitrary-schema PCD data

Browse the samples:

https://github.com/Zeniac/MPointCloud/tree/master/samples

The standalone GitHub release ZIP also includes the same sample source configured to reference the packaged MPointCloud DLL.

MEngine integration

The initial integration gate passed using the optional MEngine adapter and the published preview.2 package. MPointCloud 1.0 retains that library implementation and public API. MEngine assemblies are acquired separately.

Provider usage and mapping limits: https://github.com/Zeniac/MPointCloud/blob/master/docs/mengine-integration.md

Documentation

• Getting started: https://github.com/Zeniac/MPointCloud/blob/master/docs/getting-started.md
• Installation: https://github.com/Zeniac/MPointCloud/blob/master/docs/installation.md
• Supported formats: https://github.com/Zeniac/MPointCloud/blob/master/docs/supported-formats.md
• Performance: https://github.com/Zeniac/MPointCloud/blob/master/docs/performance.md
• PLY: https://github.com/Zeniac/MPointCloud/blob/master/docs/ply.md
• PTS: https://github.com/Zeniac/MPointCloud/blob/master/docs/pts.md

Performance

MPointCloud includes benchmarked sequential, ordered-parallel, unordered-parallel, binary, and binary-compressed I/O paths.

Performance depends on CPU core count, single-core speed, memory bandwidth, storage performance, runtime/JIT version, dataset schema/size, system load, and thermal/power limits. Published benchmark figures are comparative evidence, not universal throughput guarantees.

See:

https://github.com/Zeniac/MPointCloud/blob/master/docs/performance.md

Who uses MPointCloud?

If MPointCloud is useful in your application, we would like to hear about it. Use Contact owners on the NuGet package page to share your company/product name, an optional website link, and a short description of how you use the library.

Please state whether we may list the company/product name and link publicly. Usage reporting is voluntary and separate from required license attribution. Logos and testimonials require separate permission.

License

MPointCloud is proprietary closed-source software distributed under the MPointCloud Free Binary License.

Permitted personal, academic, research, and commercial use is available subject to the license and attribution requirements.

Copyright © 2026 Mekneo.