Painting style transformation aims to produce a painting in a particular artist's style from a picture or other paintings. In the previous implementations of exam pie-based painting style transfer, users had to ma...
详细信息
Painting style transformation aims to produce a painting in a particular artist's style from a picture or other paintings. In the previous implementations of exam pie-based painting style transfer, users had to manually select the patches from example patches. It is time-consuming and subjective for a user to select proper patches from a large number of patches. T his paper develops a patch-based approach for rendering example-based images without requiring user intervention to find appropriate patches in the synthesis process. We use mean shift segmentation and texture re-synthesis to construct an artistic database which enables users to synthesize images according to a selected painting style. Moreover, seven important painting features are proposed for finding an adequate cor respondence between the source image and the database. The synthesized output images are generated by patch-based sampling method after the correspondence has been determined. Experimental results show the feasibility of the proposed approach by demonstrating the synthesis results from different types of source images with the painting style of Vincent van Gogh.
We introduce a novel colorization framework for old black-and-white cartoons originally produced by a cel or paper based technology. In this case, the dynamic part of the scene is represented by a set of outlined homo...
详细信息
We introduce a novel colorization framework for old black-and-white cartoons originally produced by a cel or paper based technology. In this case, the dynamic part of the scene is represented by a set of outlined homogeneous regions which superimpose the static background. To reduce a large amount of manual intervention we combine unsupervised image segmentation, background reconstruction, and structural prediction. Our system allows the user to specify the brightness of applied colors unlike the most of previous approaches which operate only with hue and saturation. We also present simple but effective color modulation, composition and dust spot removal techniques able to produce color images in broadcast quality without additional user intervention. (c) 2005 Elsevier B.V. All rights reserved.
In this paper we present a novel method of hybrid-scheme for texture synthesis. By the scheme, we capture the feature of textures in two levels. At the macro-structure level, we first generate a texton mask on the pat...
详细信息
ISBN:
(纸本)0769524737
In this paper we present a novel method of hybrid-scheme for texture synthesis. By the scheme, we capture the feature of textures in two levels. At the macro-structure level, we first generate a texton mask on the pattern from the original source texture, employed in the synthesis as a primary guidance for the patch-based sampling. And in the micro-structure level, we perform blending using a pixel-based synthesis method based on the grey-level texton mask within the boundary zones of the patches. As a result, our method may not only quickly capture the macro-structure of the source texture, with little overhead from simple interactive operation,, and at the same time, the micro-structure may be also nicely preserved in the synthesis by a pixel-based blending procedure. Experiments show that our solution in combining patch-based sampling and pixel-basedsampling method based on texton masks is highly efficient for synthesizing large amount of textures with comprehensive structures.
Texture synthesis has been an active research area in computer graphics, vision, and image processing. A hybrid texture synthesis approach that combines the strength of Ashikhmin's [1] and Liang's [11] algorit...
详细信息
Texture synthesis has been an active research area in computer graphics, vision, and image processing. A hybrid texture synthesis approach that combines the strength of Ashikhmin's [1] and Liang's [11] algorithms is presented in this paper. Using patches from the input sample texture to extend the input itself, more suitable valid candidate pixels in the input sample can be more precisely determined. An extra original position array is used to record the valid pixel position in the extended region of input sample texture such that the candidate pixel can be retrieved efficiently. Pasting patches with more global information of texture features to initialize the output texture, the texture synthesis process tends to grow larger pieces in the output texture so that the perceptual similarity is reserved as much as possible. Many classes of textures have been used to test our approach. As the experimental results showed, the proposed hybrid approach effectively and efficiently synthesizes high-quality textures from a wide variety of textures.
Texture synthesis has a variety of applications in image processing. New algorithms emerged endlessly. Inspired by brushwork, we presented a novel algorithm for synthesizing textures from an input sample. First, EMD a...
详细信息
ISBN:
(纸本)0819451819
Texture synthesis has a variety of applications in image processing. New algorithms emerged endlessly. Inspired by brushwork, we presented a novel algorithm for synthesizing textures from an input sample. First, EMD algorithm was introduced and texture structure of sample was extracted. Second, we synthesized texture structure with patch-based sampling and MRF. For the spaces among structures, we extracted their irregular forms with distance transform and filled in them. Our method is not only simple and good qualitative but also robust.
We present a simple patch-based matching scheme for generating novel visual appearance in which a new image is synthesized by the optimal pasting of small patches of input sample texture image. First, we use this patc...
详细信息
ISBN:
(纸本)0780377133
We present a simple patch-based matching scheme for generating novel visual appearance in which a new image is synthesized by the optimal pasting of small patches of input sample texture image. First, we use this patch-based matching method as an efficient and simple texture synthesis algorithm to produce a wide range of textures with superb visual appearance. Second, we extend the algorithm for the texture transfer - rendering an image in the,style of a different texture image. Third, we generalize the algorithm for processing images by examples, called "patch-based image analogies". From this perspective, the texture synthesis and texture transfer are the special cases of the image texturing. Experimental results show that the proposed scheme is extremely efficient and effective for the image texturing problems, such as texture synthesis, texture transfer, tine art rendering, etc. Due to the efficiency of the patch-based sampling, our method outperforms many existing algorithms with comparable visual quality.
暂无评论