Abstract: This paper presents an example-based non-photorealistic image rendering algorithm, which employs the patch-based texture synthesis approach, as well as speeds up the texture synthesis process by exploiting particle swarm optimization during the matching search process. This algorithm also accomplishes the orientation control of matching patches in structure and details through appropriate parameter setting. Once a certain sample image with an artistic style is presented, the algorithm can accomplish image analogies by transferring the style to the target.

Key words: texture synthesis, particle swarm optimization, image analogy, artistic style learning