An optimized texture-by-numbers synthesis method and its visual applications

Abstract

The framework of texture-by-numbers (TBN) synthesizes images of global-varying patterns with intuitive user control. Previous TBN synthesis methods have difficulties in achieving high-quality synthesis results and efficiency simultaneously. This paper proposes a fast TBN synthesis method based on texture optimization, which uses global optimization to solve the controllable non-homogeneous texture synthesis problem. Our algorithm produces high quality synthesis results by combining texture optimization into TBN framework with two improvements. The initialization process is adopted to generate the initial output of the global optimization algorithm, which speeds up the algorithm’s convergence rate and enhances synthesis quality. Besides distance metrics to measure image similarities are specifically designed for different images to better match human visual perception for structural patterns and a user study is conducted to verify the effectiveness of the metrics. To further improve the synthesis speed, the algorithm is entirely implemented on GPU based on CUDA architecture. The optimized TBN method is applied to various visual applications including not only traditional TBN applications, but also image in-painting and texture-based flow visualization. The experimental results show that our method synthesizes images of higher or comparable qualities with higher efficiency than other state-of-art synthesis methods.