Detection method of fabric weft skew with camera

At present, the weft tilt angle of fabrics is mainly detected by photoelectric method or camera method. The photoelectric detection method can not adapt to thicker fabrics, and the detection accuracy is not high enough; while the camera detection method is vulnerable to local shadows, defects and noise, and can not give accurate weft tilt angle detection value of fabrics. Therefore, how to improve the accuracy of weft skew detection is a common problem faced by textile dyeing and finishing industry. In order to solve this problem, the author proposes a camera detection method for weft inclination angle of fabrics based on stepwise Otsu method and pixel statistics. Firstly, the fabric image is segmented preliminarily by Otsu method and morphological processing; secondly, the segmented image is binarized by Otsu method, and then the processed image is superimposed to form a complete binary image; finally, the accurate weft skew value is obtained according to the maximum statistical value of window pixels. Tests show that the method can still reduce the detection error to less than 0.3 degrees in the presence of shadows, defects and noises.