Research And Development Of A Portable Woven Fabric Parameters Detection System Based On Image Processing

The fabric parameters mainly consist of fabric organization,density,and unit weight.The unit weight of the fabric is obtained by calculating its weight dividing by its area,so far.While the other parameters are detected manually,mainly by some specialized equipment,such as magnifying glass and cloth mirror.The problem is that the accuracy of these detection methods is affected by the operators mood and other subjective factors,and it is also inefficient.Though,there are some textile factories trying to use industrial cameras and computers to help with the detection of fabric parameters.But it is still too complicated to apply to some complex situations.Thus we come up with a new way to help with this,which is based on image processing in a portable device to detect woven fabric parameters automatically.In this paper,a portable detection system is applied to android system of mobile phone based on digital image processing technology,which going through the mobile phone camera to obtain fabric image and using image processing software to realize woven fabric parameters automatic detection in the mobile terminal.It can improve the detection efficiency and save costs and with good flexibility by the method.To improve the accuracy of woven fabric parameters,we put forward to the following several algorithms about image processing: first of all,because of the mobile phone can not be lied in horizontal and vertical completely when the manual handhold mobile phone camera to acquire image,it will lead to distortions of the image acquisition.So the image corner detection algorithm and perspective transform algorithm by proposed for image distortion correction;secondly,because of the placement of the fabric by artificial,it will cause the fabric image tilting.The Fourier transform technology is proposed to convert the two-dimensional gray distribution of the woven fabric image into Fourier two-dimensional spectrum to analysis to realize the tilt angle detection of the woven fabric image,and correct the image;finally,we use Fourier transform technique to realize the automatic detection of fabric density and calculate area of woven fabrics based on contour extraction method.In the end,the automatic detection of woven fabric parameters is realized by applied those algorithms to the mobile phone terminal.In this paper,a portable automatic detection system of woven fabric parameters is researched and developed through applying the image processing technology to android system.The relevant algorithm program is designed to solve the key problems about the automatic detection of woven fabric parameters.Experiments show that the system has high detection accuracy,strong adaptability,wide application and good portability.The system has great practical value for the automatic detection of woven fabric parameters in textile industry.