基于图像处理的纱线检测识别系统

Study of Image Processing Based Yarn Detection and Recognition System

为实现圆纬机自动上下料系统中纱线吸取状态的判别,提出一种基于图像处理的纱线检测识别系统。该系统通过设计小尺寸盒子,与低成本相机模组、补光灯一起组成专用拍照盒。利用拍照盒内部稳定的光照环境,保证纱线图像高质量获取以及防止外部环境干扰影响识别效果。工控机对纱线吸取前后的图像进行处理,得到其像素值,最后根据两幅图像像素变化值大小,判断纱线是否成功吸取。系统搭建后,分别对8种不同颜色常用纱线进行测试,模拟头线识别情况,以及对黑色和白色纱线进行测试,模拟头线、尾线识别情况。结果表明:该方法能检测识别各类纱线吸取状态,可应用于圆纬机自动上下料系统。

To realize the discrimination of yarn in textile production,a yarn recognition system based on image processing is proposed.By designing a semi-closed cassette,together with the camera module and fill light,the system forms a special photo box for yarn identification.The stable lighting environment inside the camera box is used to ensure the high-quality acquisition of yarn image and prevent the interference of external environment from affecting the recognition effect.The obtained image is processed to obtain its pixel value,and then the change of the image is judged according to the change of the pixel value.If the change of the pixel value reaches the set value,the presence or absence of yarn can be judged.After the system is built,the single yarn identification test is carried out for 8 kinds of common yarns with different colors,and the identification test is carried out for two yarns of black and white.The results show that this method can detect and identify the absorption state of various yarns.It can be applied in the yarn breakage detection of circular weft knitting machine and the yarn capture detection of automatic joint machine.