OptiWave

Overview :

This library use optimized SVD method to perform image compression.

Features :

• Block SVD : Performs SVD using the Block Power Method.
• BPSO : Algorithm used to find optimal singular value for image compression.
• K SVD : An algorithm used to reduce the noise in an image.

Upcoming features :

• A model which works with RGB images as well.
• A model which saves image to its accurate size more efficiently.

Documentation :

For detailed information about OptiWave library, check out Documentation.

Installation :

To install the library, use the following command on terminal:

bash pip install OptiWave

Please make sure, that OptiWave and image are in the same directory.

Usage :

1. Block SVD :

python OptiWave.Block_SVD(Matrix,s)

• Block SVD function computes SVD of the input matrix 'Matrix'.
• 's' is an optional parameter, if the user wishes to compute the s'th approximation of the input matrix then 's' must be specified.

2. BPSO :

python OptiWave.svd_bpso(image,p_n,n_iterations)

• BPSO function computes the 'optimal-singular value' which is used to compress image.
• image: is a matrix which a numpy array obtained from the original image.
• p_n: It is an optional parameter, p_n need not be specified.
• n_iterations: It is an optional parameter, the user usually may not need to use it. Greater number of 'n_iterations' imply a more assured way of getting the ideal singular value which is used for image compression. Usually 3 to 4 iterations are enough to find the optimal singular value.

3. K SVD :

python OptiWave.ksvd(IMAGE : str, patch_size, stride, sparsity, max_iter)

• IMAGE: Image path.
• patch_size: Size of the square patch (pxp). It is an optional parameter.
• stride: Step size for moving the patch window. It is an optional parameter.
• sparsity: Desired sparsity level. It is an optional parameter.
• max_iter: Maxximum number of iterations. It is an optional parameter.

Contributors :

• Here is the list of the Project Members, Contributors