USING A NEW HYBRID SEGMENTATION TECHNIQUE TO SEGMENT LUNAR CRATERS AND DETERMINE ITS DIAMETER

Segmentation method is the process of partition digital image into parts depending to color, texture, and intensity. There are many segmentation methods used in different fields according to the purpose of application. In this study the global thresholding and proposed hybrid method were used to segment lunar craters. Craters on Moon's surface caused by collision between Moon and celestial objects as comet, meteorite, asteroids and others. Due to the Moon has no atmosphere, the lunar surface covered by a huge number of craters different in their size and depth depending on velocity and size of collided objects. The study of lunar craters provide information about the age and geology of a Moon's surface. So, we proposed a novel hybrid segmentation method to segment Moon's images and isolate lunar craters from other surface features and then determine the diameter of lunar craters. The proposed hybrid method combine the performance of K-Means and SFFNN together. The results shown that, the proposed method gives very acceptable outcome, where the boundaries of lunar craters were delineate in professional way that lead to accurate determination of its diameters.