Image Analysis And Measurement Of Fiber Morphology And Structure Of Nonwovens

Since the 1990 s,with the development of the textile industry,nonwovens have also been developed and applied,more and more research has been conducted based on the selection of fiber raw materials,improvements in production processes,and improvements in various properties.The structure and morphology of the nonwovens studied in this paper is of great relevance to study the properties of nonwovens.The research methods for the measurement and analysis of nonwoven fiber morphology and structure have also gradually changed from traditional manual inspection based on experience to computer-based image processing methods.The image processing methods are more scientific and objective,and have fast response and digital solutions and accuracy.This topic conducts new research on the basis of reviewing and studying the achievements of researchers at home and abroad.The sample image was first acquired using a self-built microscope acquisition system consisting of a trinocular microscope,a digital camera,a three-axis mobile platform and a computer.The prepared sample is placed on the microscope stage and adjusted in three directions of x,y,and z,so that the image can be displayed on the computer in real time,and the series of single-focus images under the focus is continuously adjusted according to a certain scale,thereby get the sample database.The resulting series of single-focus images were formed because the thickness of the nonwovens was greater than the depth of field of the microscope and the fibers could not be clearly displayed in one image.The measurement accuracy of the nonwoven fabrics depends on the clarity of the acquired image,so that the image fusion is proposed for these single-focus images,and the image fusion of such a series of single-focus images is also called multi-focus image fusion.In this paper,two multi-focus image fusion algorithms are proposed,which are GHM multi-wavelet algorithm and Non-subsampled shearlet transform algorithm.Both algorithms first decompose two single-focus images into two components,high frequency and low frequency,and then use different fusion rules to process high and low frequency components,and then combine the two processed components to obtain the inverse transform.The initial fused image is then subjected to the above steps in the initial fused image and the next single-focus image,and iteratively until all the fibers are clearly displayed to obtain the final fused image.Based on the fused image,the structure and morphology of nonwovens were measured and analyzed from three aspects: fiber diameter,porosity and fiber orientation.The measurement of fiber diameter is a combination of image preprocessing and curve fitting.The preprocessing mainly involves contrast stretching,normalization,binarization,morphology and other operations.The curve fitting is based on the idea of Hough transform.Finally,the diameter is obtained according to the distance formula,which is convenient,fast and accurate.The main step of porosity measurement is to identify the pores,calculate the area of each pore,and finally calculate the porosity.The measurement of fiber orientation is also a combination of pretreatment and curve fitting.The pretreatment mainly includes binarization,filtering and extraction of individual pores.Finally,the orientation distribution results are obtained by fitting curves with Hough transform.For the measurement of nonwoven fabric structure,this topic from collection to integration to measurement and analysis of structural parameters,the data results can be used for production guidance,promote the high-speed and automatic development of nonwoven fabric,and make the industry more continuous.