non-local-means-cuda
所属分类:GPU/显卡
开发工具:Cuda
文件大小:3293KB
下载次数:0
上传日期:2021-04-15 13:53:30
上 传 者:
sh-1993
说明: 使用CUDA进行GPU编程
(GPU programming using CUDA)
文件列表:
Makefile (591, 2021-04-15)
benchmark.sh (574, 2021-04-15)
data (0, 2021-04-15)
data\images (0, 2021-04-15)
data\images\noisy_flower.png (31899, 2021-04-15)
data\images\noisy_house.png (13726, 2021-04-15)
data\images\noisy_lena.png (81259, 2021-04-15)
data\images\original_flower.png (19288, 2021-04-15)
data\images\original_house.png (12075, 2021-04-15)
data\images\original_lena.png (45905, 2021-04-15)
data\in (0, 2021-04-15)
data\in\noisy_flower.txt (118938, 2021-04-15)
data\in\noisy_house.txt (32234, 2021-04-15)
data\in\noisy_lena.txt (506202, 2021-04-15)
data\standard (0, 2021-04-15)
data\standard\cuda_standard_5_0.06_0.8.txt (36928, 2021-04-15)
data\standard\cuda_standard_res_5_0.06_0.8.txt (38923, 2021-04-15)
data\standard\house.txt (32535, 2021-04-15)
data\standard\standard_5_0.06_0.8.txt (36928, 2021-04-15)
data\standard\standard_res_5_0.06_0.8.txt (38923, 2021-04-15)
include (0, 2021-04-15)
include\cudaFilteringGlobalMem.cuh (2677, 2021-04-15)
include\cudaFilteringSharedMem.cuh (3120, 2021-04-15)
include\filtering.cuh (1895, 2021-04-15)
include\utils.cuh (9560, 2021-04-15)
matlab (0, 2021-04-15)
matlab\add_noise.m (373, 2021-04-15)
matlab\convert_jpg.m (103, 2021-04-15)
matlab\euclideanDistance.m (177, 2021-04-15)
matlab\export_txt.m (243, 2021-04-15)
matlab\flower.mat (34321, 2021-04-15)
matlab\flower.txt (118716, 2021-04-15)
matlab\house.mat (7981, 2021-04-15)
matlab\house.txt (32535, 2021-04-15)
matlab\image.mat (14377, 2021-04-15)
matlab\lena.mat (136967, 2021-04-15)
matlab\lena.txt (505894, 2021-04-15)
... ...
# Non Local Means - CUDA
Implementation of **Non Local Means** algorithm for image denoising.
* **version 1**: CPU sequential in C++
* **version 2**: GPU parallel in CUDA
* **version 3**: GPU parallel using shared memory in CUDA
## Build instructions
**Requierements:** nvcc (Nvidia CUDA Compiler)
Set parameters in Makefile and
```bash
make
```
or
```bash
mkdir -p data/out
nvcc -o build/main -I./include src/main.cu -O3
```
and run with
```bash
./build/main $(imageNum) $(patchSize) $(filterSigma) $(patchSigma) $(useGpu) $(useSharedMem)
```
## Instructions to see the filtered images
Run the script `matlab/show_image.m` in MATLAB and all the output images of folder `data/out` will show up.
Original and noisy images can be found in `data/images`.
近期下载者:
相关文件:
收藏者: