Generally, the thermal infrared image is not easy to distinguish because of its low contrast, slow gray change, and blurred boundary between the foreground area and the background area. Fuzzy algorithm, which uses fuzzy set theory to simulate the advantages of human processing uncertain information, can well adapt to the uncertainty of thermal infrared image.

The general steps of fuzzy enhancement of thermal infrared image are as follows:

(1) Transform the image from spatial domain to fuzzy domain;

(2) Modify the features of foreground and background areas in the fuzzy domain to achieve fuzzy enhancement;

(3) The fuzzy feature is mapped back to the image plane to obtain the enhanced image.

Now take Pal-King fuzzy algorithm as an example to illustrate. First, the target area is defined as a fuzzy set. Each pixel in the image has its membership. The larger the membership value, the more it belongs to the target area. Make the membership degree of the pixel belonging to the target area μ Mn, which can convert the image from spatial domain to fuzzy domain.