Segmental Brush Synthesis with Stroke Images

Abstract

We present a new approach for synthesizing realistic brush strokes exploiting recent works of texture synthesis from stroke images. (See Figure 1). In our method, stroke images are automatically decomposed into a sequence of quad segments and stitched together along the path of user s input to produce final image. Numbers of methods using textures on digital painting have been explored; our usage of texture is novel in that the source image is typically a photo and the synthesis is fast enough to achieve realtime feedback. In contrast to previous methods, our approach allows a large variety of artistic brushes to be interactively simulated fairly so that unique media which haven t caught attention yet such as lipsticks or finger paint, are well reproduced. We shall show some artworks created using our method and demonstrate feasibility of our algorithm.We present a new approach for synthesizing realistic brush strokes exploiting recent works of texture synthesis from stroke images. (See Figure 1). In our method, stroke images are automatically decomposed into a sequence of quad segments and stitched together along the path of user s input to produce final image. Numbers of methods using textures on digital painting have been explored; our usage of texture is novel in that the source image is typically a photo and the synthesis is fast enough to achieve realtime feedback. In contrast to previous methods, our approach allows a large variety of artistic brushes to be interactively simulated fairly so that unique media which haven t caught attention yet such as lipsticks or finger paint, are well reproduced. We shall show some artworks created using our method and demonstrate feasibility of our algorithm.We present a new approach for synthesizing realistic brush strokes exploiting recent works of texture synthesis from stroke images. (See Figure 1). In our method, stroke images are automatically decomposed into a sequence of quad segments and stitched together along the path of user s input to produce final image. Numbers of methods using textures on digital painting have been explored; our usage of texture is novel in that the source image is typically a photo and the synthesis is fast enough to achieve realtime feedback. In contrast to previous methods, our approach allows a large variety of artistic brushes to be interactively simulated fairly so that unique media which haven t caught attention yet such as lipsticks or finger paint, are well reproduced. We shall show some artworks created using our method and demonstrate feasibility of our algorithm.