Method of spatial domain processing in image processing

In order to avoid the text expression is not intuitive, directly above to illustrate the spatial domain processing method.

Here are some of the methods mentioned above. Numeric Operations

As shown in the following diagram, the numerical operation can be divided into single-channel image processing and multi-channel image processing because of the specific Operation object.

Single-channel point operation is the direct processing of the value of pixels, such as logarithmic transformation, power-law transformation, piecewise transformation, and so on is a mathematical formula applied to each pixel point. And the neighborhood operation is a typical filter, such as mean smoothing, Gaussian smoothing, Sobel filter and so on. Set Operation

The basic operation of the same image and the difference between the set, intersection and difference of different images.

For example, the image mosaic is looking for two images of the set (overlapping area) stacked together, and then formed a new image, intuitively, the image becomes larger. The following figure:
Logical Operations

The logical operation is relatively simple, because it is a logical matrix, only 0, 12 cases. Basically the same as our previous logical operations, but this is mostly the form of matrices.
Mathematical Morphology manipulation