Research On Fabric Unidirectional Moisture Conductivity Detection Method Based On Image Processing

Unidirectional moisture conducting fabric is a new functional fabric developed on the basis of moisture absorption and quick drying.It mainly uses the difference of water absorption between the inner and outer layers of the fabric to achieve the effect of unidirectional transmission of water in the fabric.At present,the method of measuring droplet area is complex and subjective;Some scholars also applied image processing technology to their measurement,and obtained the wetting area of the fabric at each time quickly and accurately through the continuous improvement of morphological processing and segmentation methods.On the basis of predecessors,through the introduction of moisture conduction theory and fluid equation,this topic explores the functional relationship between fabric diffusion rate and water absorption,and improves the disadvantage of single detection index of traditional image method.It mainly involves:(1)Acquisition of wetting image: this subject designs an experimental device,which uses a precision catheter to control the flow rate and total amount of injected liquid,and places cameras on and under the support table of the device to obtain the images of the inner and outer layers of the fabric at the same time.Finally,the images are transmitted to the computer by taking frames at a fixed time for processing.(2)Measurement of droplet contour area: after the computer receives the image group,the image is processed successively by morphology and image segmentation to obtain the fabric contour.The wetting area and the actual size of the radius are calculated quantitatively through the pixels,and compared with the actual results by the weighing method.The results show that the error between the area measured by the image method and the actual area is within 5%,but the speed is faster.(3)Calculation of fabric moisture conductivity index: the first is the measurement of water diffusion rate,which is to obtain the actual wetting radius of droplets,draw the radius time curve,and take the cumulative diffusion rate of water as its evaluation index;Then,by simplifying the theory of fabric moisture conductivity and limiting the experimental conditions,the hydrodynamic equation is introduced to explore the functional relationship between fabric water absorption rate and diffusion rate.After verification,the functional relationship is reliable and the measurement of fabric water absorption rate is realized;Finally,based on the former,through the integral difference of the water absorption rate time curve of the fabric,the water content difference between the inner and outer layers of the fabric is calculated to obtain the unidirectional moisture conductivity of the fabric.(4)Rating training of fabric moisture conductivity: use liquid moisture management tester(MMT)to detect the above properties of fabric,and analyze the correlation between the data results measured by the two methods.Referring to MMT,the performance indexes measured by image method are taken as the input function,and the moisture conductivity grade evaluation of fabric made by MMT is taken as the output function,which is put into the rating training of BP neural network.The results show that the accuracy of this method for the fabric training results of grade 2-4 moisture conductivity grade is high,which can reach more than 95%,but the training results of grade 1 and grade 5 fabrics are poor,which needs further improvement.