Abstract: In this paper, a new method for the synthesis of stroke patterns is proposed, which considers both local and global features of an input stoke pattern.The proposed method takes a small pattern sketched by the user as an input sample.First the representative elements in the pattern image are extracted from the sample and clustered into classes.Then the local and global features of the sample are analyzed.Finally the clustered elements extracted from the sample are arranged based on these features to form large patterns perceptually similar to the input small sample.The experimental results show that the patterns synthesized by the proposed method inherit the style of the input sample and do not simply repeat the small sample.Compared with the related methods, the proposed method well solves the problem by utilizing both local or global features in the synthesis of stroke patterns.