iLab Neuromorphic Robotics Toolkit  0.1
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Macros Groups Pages
Low-Pass / Blur Operations

Methods for performing low-pass (blurring) operations on Images. Each of these methods takes a filter size (N), which they will use to select a proper filter at runtime. If N is 3, 5, or 9, then these methods will just forward the work to some very optimized methods (see Low Pass Helpers ). Otherwise, a new binomial kernel will be computed on the fly and the Image will be filtered with the newly computed kernel. Thus, if you need to perform a low pass many times with some odd-ball sized kernel, you could save a few clock cycles by computing the kernel manually using the binomialKernel() method, and then passing it to separableFilter().

See Also
Image Filtering

Modules

 Low-Pass Helpers
 

Functions

template<NRT_PROMOTE_PIX_NO_DEFAULT_PROMO >
Image< DestType > nrt::lowPass (int const N, Image< PixType > const &src)
 Low-pass filter, NxN applied in X and Y. More...
 
template<NRT_PROMOTE_PIX_NO_DEFAULT_PROMO >
Image< DestType > nrt::lowPassX (int const N, Image< PixType > const &src)
 Low-pass filter, Nx1 separable, applied in X.
 
template<NRT_PROMOTE_PIX_NO_DEFAULT_PROMO >
Image< DestType > nrt::lowPassY (int const N, Image< PixType > const &src)
 Low-pass filter, 1xN separable, applied in Y.
 

Function Documentation

template<NRT_PROMOTE_PIX_NO_DEFAULT_PROMO >
Image<DestType> nrt::lowPass ( int const  N,
Image< PixType > const &  src 
)

Low-pass filter, NxN applied in X and Y.

This method just calls lowPassX() and lowPassY() in sequence.