Study On Fiber Orientation Of Nonwovens Based On Image Fusion

Different from traditional textiles fabricated with yarns in certain ways, nonwoven material is directly composed of fibers. Fiber is the basic raw material of the nonwovens, its properties exert direct influence on the nonwovens’performance. As an important characteristic of nonwovens, fiber orientation distribution in webs directly influences the performance of material, especially its mechanical anisotropy behavior. What’s more, manufacturing process and application field of nonwovens are partially determined by the arrangement of fiber in the assembly. Therefore, it’s very necessary and important to obtain accurate fiber orientation. At present, there are many ways to measure fiber orientation in nonwovens, mainly include direct and indirect methods. Direct ways are primarily original tension, microscope observation etc., indirect methods include mechanical performance anisotropy, microwave and laser scattering and so forth. In spite of the fact that these methods are reliable, however, they are time-consuming and laborious and can not obtain fiber orientation structurally.In recent years, with the popular application of image processing technology in textile industry, many researchers have proposed the method using image processing to recognize and decide fiber orientation in the assembly. However, so far. this technique can just been used to acquire clear images for thin nonwovens (usually less than10g/m2). For those whose thicknesses go beyond the depth-of-field of microscope, it fails to capture sharply-focused images. At present, SEM and CT technologies are usually used to acquire clear images. These devices, however, are expensive, slowly and demanding on the operation. These limitations and shortcomings restrict its wide applications in fiber orientation analysis. Based on these problems, the following solutions were put forward in the current paper.This paper adopted image fusion technique to acquire sharply-focused images in super depth-of-field for nonwovens to solve problems such as image blurring and serious information loss. Automatic focusing system was used to take a series of multifocus images at one scene, and then image fusion technique was applied to combine them into one clear image.In the wake of images being subjected to several steps of image processing, discontinuous fiber segments occur, one algorithm called boundary curve integral was proposed to calculate tangent angle for fibers’ two edge lines. Then the fiber orientation distributions in the comprehensive range of0to180degree were obtained.This paper also presented manual-tracking for SEMs and short-span tensile test at intervals of15degrees to verify the usefulness of boundary integral algorithm in getting the fiber segments orientation in fused images. The experimental results demonstrate that the proposed method for fiber orientation is quick and accurate. Meanwhile, this paper classified fiber orientation into two categories:Pure Fiber Orientation (PO) and Enforced Fibers Orientation (EO).In this paper, image fusion technique was used to capture sharply-focused images for nonwovens in the super depth-of-field. The boundary curve integral algorithm was applied to calculate fiber segment’ orientation, and then fiber orientation distributions within a comprehensive range of0to180degree were realized. Moreover, this method is expected to replace the existing methods for fiber orientation.