MimWaveletDenoise

Specifies the identifier of a wavelet image processing context, when performing a wavelet denoising of an image. The context must have been allocated using MimAlloc() with either M_WAVELET_TRANSFORM_CONTEXT or M_WAVELET_TRANSFORM_CUSTOM_CONTEXT. When performing a wavelet denoising of a result, this information is not required; set this parameter to M_NULL.

If you specify a wavelet context, it indicates the filter and mode that this function uses to perform the wavelet denoising. For an M_WAVELET_TRANSFORM_CONTEXT context type, you can set a predefined type of wavelet filter, using MimControl() with M_WAVELET_TYPE. The default is a Haar wavelet filter type. For an M_WAVELET_TRANSFORM_CUSTOM_CONTEXT context type, you must specify custom wavelet filters, using MimWaveletSetFilter(). For either type of context, the default mode is dyadic. To change this (for example, to specify undecimated), use the M_TRANSFORMATION_MODE control.

To inquire about the information in a wavelet image processing context, call MimInquire(). For example, you can retrieve the identifier of the internal buffer containing the wavelet filter values (M_FILTER_..._ID).

Specifies the identifier of the source with which to perform the wavelet denoising. The identifier must be for an image buffer or a wavelet image processing result buffer. The image buffer can be of any type. The wavelet result buffer must have been allocated on the required system using MimAllocResult() with M_WAVELET_TRANSFORM_RESULT. It must hold MimWaveletTransform() results. To retrieve information about wavelet transformation results, use MimGetResult() and MimGetResultSingle().

If the source is an image buffer, the destination must be an image buffer of the same type. If the source is a wavelet result buffer, the destination must be a wavelet result buffer.

If you specify an image buffer, there must not be a region of interest (ROI) associated with it. Using an image buffer with an ROI will cause an error.

Specifies the identifier of the destination with which to perform the wavelet denoising. The identifier must be for an image buffer or a wavelet image processing result buffer. The image buffer can be of any type. The wavelet result buffer must have been allocated on the required system using MimAllocResult() with M_WAVELET_TRANSFORM_RESULT.

If the destination is an image buffer, the source must be an image buffer of the same type. If the destination is a wavelet result buffer, the source must be a wavelet result buffer.

If you specify an image buffer, there must not be a region of interest (ROI) associated with it. Using an image buffer with an ROI will cause an error.

Specifies the transformation level at which to perform the denoising.

Set this parameter to one of the following:

For specifying the transformation level
Value	Description
M_DEFAULT	Specifies that the level information is not required. Only valid when the source is a result. Denoising is performed at the transformation level indicated in the result. (summarize) Specifies that the level information is not required. (more details...)
0 < Value <= MaxLevel	Specifies the level. Only valid when the source is an image. This function decomposes the image according to the specified number of levels, performs the denoising, and recomposes the image using the same number of levels. MIL establishes the maximum level to be the point at which denoising at subsequent levels does not provide relevant data. The maximum level depends on the source image and the wavelet image processing context. If the specified level exceeds the maximum level, MIL uses the maximum level. (summarize) Specifies the level. (more details...)

Specifies the type of wavelet process with which to perform the denoising.

Set this parameter to one of the following:

For specifying the wavelet shrinkage process
Value	Description
M_BAYES_SHRINK	Specifies predefined BayesShrink denoising techniques. Typically effective for minimizing Gaussian noise. (summarize) Specifies predefined BayesShrink denoising techniques. (more details...)
M_NEIGH_SHRINK	Specifies predefined NeighShrink denoising techniques. Typically effective when noise is difficult to characterize. Denoising mimics the general behavior of wavelet coefficients by taking neighborhood statistics and dismissing outliers. M_NEIGH_SHRINK is often faster than the other predefined denoising techniques, though results can be less precise. (summarize) Specifies predefined NeighShrink denoising techniques. (more details...)
M_SURE_SHRINK	Specifies predefined SureShrink denoising techniques. Typically effective for minimizing Mean Square Errors (MSE). (summarize) Specifies predefined SureShrink denoising techniques. (more details...)

Reserved for future expansion and must be set to M_DEFAULT.