An Image Generation Method for Enbroidery

 In recent years, machine embroidery enables a semi-automatic mass production of fine embroideries. However, the input data for machine embroidery has been generated manually, and this manual process requires a lot of time and effort. Therefore, we propose an image-based embroidery rendering to semi-automate the manual process for cost reduction.  Our method is based on real-world constraints of machine embroidery to generate embroidery-like image. First, we input the image. Second, embroidery colors are selected by color reduction and approximation by the database generated from real embroidery colors. There are constraints of the mechanical limit of the number of embroidery colors and the material limit of the available color. Then, we render the embroidery image by texture mapping of embroidery threads. We calculate parameters of texture mapping such as position, size, orientation, order, and color. Here, we diffuse errors of by much fewer colors than in the input image.