Abstract: In this paper, a ship texture automatically extracted from high resolution remote sensing images is researched. The purpose of study is to manage some specific ships in Changjiang River according to requirements of relative departments, and monitor types, distribution and quantity of ships in a specific time and a specific region. Adopting indexes in statistics is selected as estimate factors. Application to ship texture has been realized by IDL(interactive data language), a powerful tool in process of images. By analyzing visual interpretation and results of experiment, it is shown that selected index is reasonable. Entropy, gradient and relativity between them, are the most appropriate texture factors to represent features of the ship texture.