Faint targets detection based on gray morphological filtering and neighborhood entropy method

A new method of faint-target detection is put forward. To improve the practicality of the method, especially it is used to deal with those images that involves the complex background and a lot of noise, the method of neighborhood entropy was brought forward. Furthermore, to increase the validity of it, this method was combined with gray morphological filtering. First ,the image is processed with morphological filtering. Second, the neigborhood entropy is calculated. And then, the target is detected according to the maximum value and minimum value of entropy. The experiment results indicated the new method can detect small target stably and guickly.