Fuzzy region competition-based auto-color-theme design for textile images

摘要: We propose a model to recolor textile images by different color themes. The contains three phases. first phase is partition an input image into several homogeneous regions. CIELab mean of each region and bias-field function are obtained from the segmentation results. combination values all regions considered as theme image. second retrieve relevant themes given dataset. retrieved preserve mood in sense similarity measurement defined space. In third phase, we reconstruct new with appearances using proposed method provides powerful tool for designers generate search combinations related theme. Numerical results indicate that our recolorization performs well on complex design patterns.