https://github.com/saronic-technologies/libdebayer
https://github.com/saronic-technologies/libdebayer
Last synced: 9 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/saronic-technologies/libdebayer
- Owner: saronic-technologies
- License: mpl-2.0
- Created: 2024-09-05T00:33:24.000Z (almost 2 years ago)
- Default Branch: main
- Last Pushed: 2024-10-22T01:17:20.000Z (over 1 year ago)
- Last Synced: 2024-10-23T00:54:20.000Z (over 1 year ago)
- Language: C++
- Size: 3.39 MB
- Stars: 2
- Watchers: 3
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# libdebayer
A CUDA accelerated debayering library with C/C++ and Rust support.
## Dev Setup
```
nix develop
```
This enters a dev shell with all of the necessary build dependencies.
## Library Breakdown
### libdebayer
This is a simple C API that wraps the underlying CUDA kernels. The C
API assumes the images are already in GPU memory. libdebayer
implements 3 debayering algorithms:
- [Menon 2007](https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=8c8e4a3cf6d0b8dfdcd36652718ad54afd2fe5fe)
- [Malvar 2004](https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/Demosaicing_ICASSP04.pdf)
- Bilinear (same as builtin OpenCV)
### libdebayercpp
This provides a higher level C++ API that performs cudaMemcpy's from
host to device. An example of how to use this library is in
`benchmark/cpp`.
### libdebayer-rs
This provides a higher level Rust API with `TryFrom` traits to go from
an OpenCV `Mat` to a debayered `Mat`. An example of how to use this
library is in `rust/examples/test_benchmark.rs`. This test benchmark
is identical to the `benchmark/cpp` program.
### Run Rust Example
```
nix develop
cd rust
KODAK_FOLDER_PATH="" cargo run --example test_benchmark
```
## Benchmark
libdebayer is benchmarked against OpenCV and NPP (Nvidia Performance
Primitives). To run the benchmark for the Malvar 2004 implementation
run the following: `nix run .#kodak_benchmark_cpp`
### Benchmark Results
- OCV-EA gets Average PSNR: 28.616 dB
- NPP gets Average PSNR: 28.8522 dB
- libdebayer Malvar 2004 CUDA kernel gets Average PSNR: 33.4554 dB (+4.8394 dB)
- libdebayer Menon 2007 CUDA kernel gets Average PSNR: 37.6907 dB (+9.0747 dB)
- Eliminates color fringing
#### Image Results
Example output from OpenCV-EA:
Example output from NPP (Nvidia Performance Primitives):
Example output from our Malvar 2004 CUDA kernel:
Example outputs from our Menon 2007 CUDA kernel:
