Frequency Domain Image Enhancement Technology

1. In an image, the gray value of the pixel changes frequently with the position, which can be expressed by the frequency. This is a spatial frequency that varies with the position.

It refers to the difference between the maximum value and the lowest value of the continuous pixel gray value. Jenkins defines the spatial frequency as "the number of changes to the brightness value per unit distance of a specific part of the image ".

On the frequency domain plane, the low frequency zone is located at the center and the high frequency zone is located at the edge.

2. Filtering

Filtering is an improved output image technology that applies certain filtering functions to the input image in the image space (x, y) or frequency (X ', y.

(1) spatial domain filtering

For digital images, spatial domain filtering is performed by local products and operations (also called convolution). Generally, nxn matrix operators are used as convolution functions.

(2) frequency domain filtering

Frequency domain filtering is represented by the product of Fourier transformation.

====================================== Common spatial domain filtering convolution functions + frequency domain filtering functions ==== ============================