The automatic segmentation of flaws and acquisition of parameters in woven fabrics has been achieved. However, the problem is that information of gray information of float is difference with real fabric. It is important to use a fabric that includes the total information of woven fabrics, meanwhile, the fabric was illuminated even. To solve the problem, in this research, images of virtual woven fabrics of plain were developed in order to research the methods of automatic segmentation of flaws and acquisition of parameters in woven fabrics using the images of real woven fabric. The warp and weft float in the fabric were cut from the scan image of real fabric. The fabrics included main information of interlacing of warp and weft, the count, weaving density, and reflection property of real float. It was possible that the fabrics could be used to discuss the problem proving robustness of the algorithms by adding unique flaw to the fabric.