MATLAB CVIP Toolbox Functions
Spatial Filters: Adaptive Filters
The spatial filter group in the CVIP Toolbox has a variety of different types of functions. They are organized into the separate categories of adaptive filters, mean filters, order filters and a miscellaneous group. The adaptive filters category contains spatial domain based filters that are used primarily for contrast enhancement, noise mitigation or image smoothing. The adaptive filters included here function by adjusting parameters based on the characteristics of the underlying local pixel values. The characteristics can be standard statistics, order statistics or neighborhood image features, such as edge energy. Included are filters to adjust image contrast, such as the ACE (adaptive contrast enhancement) filters. Filters for noise removal, such as the adaptive median and minimum-mean-squared-error filters are also part of this group.
The mean filters are primarily smoothing filters, which perform various methods of averaging and will provide noise mitigation or simply add a “soft” look to the image. The order filers include the median, midpoint, minimum, maximum, and the alpha-trimmed mean which varies between a mean and a median filter to handle multiple types of noise. Utility functions for convolution, which allows the implementation of any linear spatial filter mask, and difference filters are included in the miscellaneous group.
- ace2_filter_cvip - adaptive contrast and enhancement filter
- adaptive_contrast_cvip - adaptive contrast filter, adapts to local gray level statistics
- adapt_median_filter_cvip - a ranked-order based adaptive median filter
- ad_filter_cvip - anisotropic diffusion filter
- exp_ace_filter_cvip - performs an exponential ACE filter
- improved_mmse_filter_cvip - perform improved adaptive minimum mean squared error filter
- kuwahara_filter_cvip - an edge preserving, smoothing filter
- log_ace_filter_cvip - performs a log ACE filter operation
- mmse_filter_cvip - minimum mean squared error restoration filter