An Object-oriented Fuzzy Enhancement for Remote Sensing Images

Image enhancement is one of the important image processing techniques for remote-sensing image.It is true that there exists a lot of methods of image enhancement,but most of them are heuristic and problem dependent.Each of them is just suitable for a given problem and purpose of application.In this paper,considering the ambiguity and uncertainty in grayness of image containing an object,an object-oriented method of fuzzy enhancement for remote sensing image is presented to well highlight an interested object of image and to restraint the background.Unlike the common method of fuzzy enhancement of image,a trapezoid-like fuzzy member-ship function is designed to transform the intensity domain to the fuzzy domain.In addition,it is not always necessary that the fuzzy membership of a crossover point must be equal to 0.5 in our method.This technique can not only highlight the interested object according to one's requirements and the feature of image but also make the blur edge of object clearer.The experiments show that the results are satisfactory.