Density detection for knitted fabric based on image space domain method

In textile processing, the density of knitted fabric is mostly measured offline by using weft mirrors, which is time-consuming and discontinuous. Meanwhile, offline measurement cannot meet the requirements for online control of the processing. An image based method is proposed in this paper to deal with the online detection of the density of knitted fabric. First, by binarizing and filtering the image of fabric, the clear texture of the fabric can be achieved. After extracting the yarns based on the feature of the texture, the Hough transform is used to recognize the extracted yarns. Then the horizontal density of the fabric can be calculated. Next, through analyzing the geometric features of the fabric´s coils, the mean distance between each two coils can be obtained. Then the vertical density of the fabric can also be calculated. From the horizontal and the vertical density, the general density of the detected fabric is obtained easily. At last, the results of experiments demonstrate the effectiveness of the proposed method.