Style Translation And Protection Of Copyright For Line Drawings

Style learning and transformation have been a hotspot in NPR area. Art designs (line drawings)are widely used in scientific literature illustrations, graphic design industrial design and plane design, etc. So that, style re-using and transformation have been especially important. On the side of patterns design and marking, some enterprises such as printing, dyeing and ceramics, advertisement, cartoon marking, etc, behave poorly in innovation and manufacturing because of difficulty in re-using and adjusting the style. Up to now, researchers have studied and found many effective algorithms for style translation, however, these algorithms are still not perfect in theory. Therefore, the further research on methods of artistic styles not only conducive the improvement of algorithm theory but also help promote the method of application.With the development of computer network technology, all kinds of multimedia works such as audio, video, animation, images and so on have released in the form of the network. Violators can copy and disseminate the unauthorized digital images at the lower cost. This has also become an important factor restricting the development of enterprises. So the line drawing on the way for copyright protection was proposed.This paper is focused on technologies for a more systematic analysis and research, and a method of style translation for line drawings based on example regional lines was presented. Based on the analysis and conclusion of the digital watermarking algorithm for binary image, an adaptive binary image watermark algorithm based on the combination of space and frequency domain was proposed. Include the following elements:(1) NPR overview and line drawings translation on the basis of knowledge. First the concept of non-photorealistic rending, the background and related research are introduced. Second, the paper introduces the line-drawings features, analyzes the line drawings method which computer generated, and also proposes a brief overview of the style translation methods.(2) Prior knowledge of style translation for line drawings. The style of a line drawing is abstracted as the styles of its elements: strokes, which are the trace of the artists'pen. With standard representation of the strokes'shape and width, the methods of style translation for line drawings are discussed. And the extraction of strokes, drawing of strokes and the methods for similar comparison are also introduced in this paper.(3) Style translation for line drawings based on example regions. The method of style translation for line drawings based on example can be used to obtain the image with the target image style, and it seems example as a whole style. However, some line drawings have included a variety of styles of line, and different styles of lines can give the different impression. On the basis of the above, an improved method of style translation for line drawing is proposed. The experiments showed that this method could be more flexible in the style translation for line drawing.(4) A binary image zero-watermark algorithm based on the DCT domain. First this paper introduces the characteristics of the binary image and summarizes the binary image watermark algorithm. Finally, a binary image non-watermarking algorithm based on the DCT domain is proposed, because the DCT coefficient (low frequency) sign basically unchanged, when images receive a variety of geometric attack. Simulation shows that the method has the better anti-geometric attack capability, robust, and has good practical value.