The previous section we introduced the image frequency domain transformation, this section will be based, describes the content image frequency domain filtering comprising a common high-pass / low-pass / band-pass / band-stop / direction filtering frequency domain filtering methods, while coding in C language, help beginners to understand and grasp how the frequency domain filtering images.
[Definition and Algorithm]
The image frequency domain filtering means to convert the image from the spatial domain to the frequency domain through Fourier transformation, the frequency spectrum diagram of an image filtering, and then the restored image information is filtered to the filtering method of the spatial domain by the inverse Fourier transform . It can solve some weaker characteristic in the spatial domain, the image signal processing but significant problem in the frequency domain characteristics. Frequency domain filtering also comprises high pass filtering, low pass filtering, bandpass filtering, and filtering and much direction. As a beginner, we want to have is frequency domain filtering before giving top priority, and apply their knowledge.
Frequency domain filtering process follows the general method shown in FIG Fig.1:
For an image S, Fourier transformed spectrum F resulting FFT, using a filter F for filtering of t, that is multiplied by the new spectrum F ', and then subjected to inverse Fourier transform the IFFT, get results in Figure D. T for selective filter, depending on the type of filter effect will be different, Here we introduce some common filter.
Low Pass Filter: Let the low frequency information, filtering high frequency information. Low-pass filtering over its template as follows:
