evan/OpenDepthMap
Archived
1
Fork 0
Code

add in missing links

Browse Source
This commit is contained in:
Evan Pratten 2021-05-29 20:37:51 -04:00
parent 85ed858301
commit ed61f4a72b
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
README.md
View File

@ -12,9 +12,9 @@ OpenDepthMap is split into three parts, across four distinct programming languag
LeapMotion Driver <---> LeapMotion Library <---> FFI Wrapper <---> Rust Interface Library <---> CV Pipeline LeapMotion Driver <---> LeapMotion Library <---> FFI Wrapper <---> Rust Interface Library <---> CV Pipeline
``` ```
The LeapMotion devices can only interact with Linux hosts via the [`LeapMotion V2`]() API. This API is written in C++ and has no C bindings, making it hard to expose via Foreign Function Interface to another language. I solved this problem by writing a [small wrapper]() that exposes the C++ functions I need via a C interface. A Rust library ([`libodm`]()) then uses this wrapper to expose a type-safe, memory-safe, and error-handled interface to the LeapMotion device in a format that is easy to generate further FFI bindings from. The LeapMotion devices can only interact with Linux hosts via the [`LeapMotion V2`](https://developer-archive.leapmotion.com/documentation/v2/cpp/index.html) API. This API is written in C++ and has no C bindings, making it hard to expose via Foreign Function Interface to another language. I solved this problem by writing a [small wrapper](./libodm/cpp/wrapper.cc) that exposes the C++ functions I need via a C interface. A Rust library ([`libodm`](./libodm)) then uses this wrapper to expose a type-safe, memory-safe, and error-handled interface to the LeapMotion device in a format that is easy to generate further FFI bindings from.
Originally, this project ended at Rust, and made use of the [`opencv`]() crate to handle the computer vision pipeline, but I was limited by a few critical functions missing from that binding, so I decided to add yet another layer of FFI into the mix. Using [PYO3](), `libodm` is exposed as a Python3 library, where the end-application uses it to control and read from the LeapMotion device, while also using Python's OpenCV bindings and [`numpy`]() for the final processing step. Originally, this project ended at Rust, and made use of the [`opencv`](https://crates.io/crates/opencv) crate to handle the computer vision pipeline, but I was limited by a few critical functions missing from that binding, so I decided to add yet another layer of FFI into the mix. Using [PyO3](https://github.com/PyO3/pyo3), `libodm` is exposed as a Python3 library, where the end-application uses it to control and read from the LeapMotion device, while also using Python's OpenCV bindings and [`numpy`](https://numpy.org/) for the final processing step.
## Vision Pipeline ## Vision Pipeline