TV Image Denoising CUDA
TV Image Denoising CUDA is a command-line tool and graphical user interface (GUI) program designed for denoising images using the Total Variation (TV) denoising algorithm. It leverages the power of CUDA, a parallel computing platform and application programming interface (API) model created by NVIDIA, to accelerate the denoising process on compatible GPUs.
Features
- Denoises images using the TV denoising algorithm.
- Supports various image formats.
- Allows customization of denoising parameters:
- Lambda: The regularization parameter for TV denoising.
- Iterations: The number of iterations for TV denoising.
- Provides a user-friendly GUI for easy image selection and visualization.
- Saves the denoised image to a specified output file.
Usage
To use TV Image Denoising CUDA, follow these steps:
Build the program using the provided source code and necessary dependencies.
Run the program with the following command:
./TV_Denoising_CUDA <input_image> <output_image> <lambda> <iterations>
- input_image : The path to the input image file you want to denoise.
- output_image : The path to the output denoised image file.
- lambda : (optional): The regularization parameter for TV denoising (default: 0.02).
- iterations : (optional): The number of iterations for TV denoising (default: 10).
The program will open a GUI window displaying the original image and the result image. You can interact with the GUI to adjust the denoising parameters and visualize the denoised image in real-time.
Click the "Denoise" button to perform TV denoising on the selected image using the chosen parameters.
The denoised image will be displayed in the GUI, and the program will save it to the specified output file.
Dependencies
- Qt framework (Qt GUI library)
- OpenCV library
- CUDA toolkit (compatible GPU required)