mcmc-denoising

KO | EN

gitlite — search

#python #react #ai #llm #android #machine-learning #linux #docker #nodejs #java #awesome #vue #pytorch #openai

(★ 8)

Bayesian image denoising in C++ and Python

mcmc-denoising 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Bayesian image denoising

This project uses Markov-chain Monte Carlo (MCMC) methods to sample from a prior distribution of images in order to perform denoising on images. The algorithms perform well on images with *speckle* noise as well as *salt & pepper* noise.

This project is built up of two C++ classes, `Ising` and `Potts` for binary and grayscale images respectively. Additionally, we provide a Cython bridge that enables use directly from Python.

## Examples

![Ultrasound](https://raw.githubusercontent.com/wohlert/mcmc-denoising/master/data/fig-ultrasound.png)

![SAR](https://raw.githubusercontent.com/wohlert/mcmc-denoising/master/data/fig-sar.png)

## How to use

In order to use the library from Python you will need a C++ compiler along with an edition of Cython. Compile the Cython library by running the following.

```bash
python setup.py build_ext --inplace
```

This will generate the library, typically `denoising.cpython-*.so` which you can import from Python.

Now it is possible to import the library. The example below shows how to generate the image in the ultrasound example.

```python
import numpy as np
import matplotlib.pyplot as plt

from denoising import PottsMH, MPM

image = plt.imread("data/ultrasound.jpg")

potts = PottsMH(image, beta=4, sigma=np.std(image), bins=15)
denoised = MPM(potts.solve, iterations=200, images=20)

plt.imsave("ultrasound-clean.png", denoised)
```

mcmc-denoising

mcmc-denoising 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

mcmc-denoising

mcmc-denoising 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README