Development And Application Of Intelligent Yarn Detection System Based On Digital Image Processing Technology

In domestic, textile industry as the traditional industry, is the advantage industry, havemade great contribution to national economy. However, our textile industry fell behindwith poor degree of information from a global point of view. The present yarn detectionmeans existence problem of poor reliability and low sensitivity in the textile industry.Therefore, a digital intelligent detection mean was proposed to determine accurately theyarn more, less, especially breaking in the process of spinning yarn, which is formed by theprecise CCD imaging sensor in conjunction with the mature digital picture processingtechnique.The intelligent yarn detection mean dynamically capture the high resolution image vialinear CCD and analyze image data by image processing module. The communicationcontrol module in the system would stop the operation of warper and show the faultlocation when the yarn more, less, or breaking via comparation the quantity of spinningyarn with the standard quantity of yarns. The lateral resolution and the horizontal pixel ofyarn image captured by linear CCD respectively are4800dpi and20400. The image datawould be processed by image processing module, such as graying, denoising and contraststretch. The key algorithm is to analyze all of pixels with one line high frequency, to selectand correct maximum value by second order difference, to segment threshold byone-dimensional Otsu adaptive method. Then the singlechip control unit would stop theoperation of warper when the maximum value was unequal to the standard quantity ofyarns. Thus, the system reach the goal of automated testing.