A method of point target identification based on fuzzy set theory

There are many objects and noises in an infrared image with complex sky-sea background, some of these objects and noises exist as points in an image. When a warship appears as some points in an infrared image, it is difficult to identify which point is a target (warship). By analysis these images, this paper presents a method used Fuzzy degree of nearness to identify the target from many objects and noises in an image. At first, based on the conditions recognizing target, we define three Fuzzy sets for targets: (1) the largest distance between sky-sea line target; (2) the area of a target; (3) the maximum distance between two chimneys of a target. Then we gather together these three Fuzzy sets to form a general Fuzzy sets for a target. At last, by comparing general Fuzzy sets for a target with the one for an uncertain point in an image, the Fuzzy degree of nearness for an uncertain point is obtained. By calculating all the Fuzzy degree of nearness for all points in an image, the maximum Fuzzy degree of nearness for a certain point can be found, thus this point is identified as the target. The algorithm based on this method is verified by SO infrared images, it has high robust suitability for identifying the point target in the infrared images with complex sky-sea background.