Font Size: a A A

Study On Objective Assessment System Of Fabric Smoothness Grade

Posted on:2007-05-25Degree:MasterType:Thesis
Country:ChinaCandidate:J M XuFull Text:PDF
GTID:2121360182978436Subject:Textile Engineering
Abstract/Summary:PDF Full Text Request
Fabric appearance is one of the most important properties of textile quality control and fabric smoothness is the attribute of textile appearance evenness. Traditionally, fabric smoothness is evaluated by comparing the sample with the standard replicas through eyes, which is subjective and poorly precise. Computerized image processing techniques are now widely used in the inspection and quality control of textile industry, which can carry out transformation on the image and extract characteristic information from the image conveniently. Therefore it is a trend of using image processing techniques to evaluate the fabric appearance objectively.In this thesis, the objective assessment system of fabric smoothness grade is developed based on photometric stereo method. Fabric wrinkle presents a 3-D zigzag appearance, but it is just a 2-D gray image that captured by CCD camera. The idea of the photometric stereo method is to vary the direction of incident illumination between successive images while holding the viewing direction constant. The 3-D surface shape of fabric can be reconstructed based on these 2-D gray images, which provides sufficient information to study surface appearance properties of the fabric.In this thesis, image-capturing device is designed and made based on photometric stereo method to capture the fabric images. Then the 3-D surface information of fabric is obtained in the light of the reconstruction of the fabric surface shape based on photometric stereo method. The feature parameters that can reflect fabric smoothness degree are extracted by using the acquired surfaceinformation. Finally, taking these feature parameters as the input parameters of BP nerve network, the nerve network will evaluate fabric smoothness grade. The main conclusions of this thesis are briefly summarized as follows:(1) Based on photometric stereo method, the objective assessment system of fabric smoothness grade has been developed, which integrates the image process and analysis program with the image-capturing device.(2) The image-capturing device has been designed and made. This device is not complicated, but convenient to capture the fabric image fast. As the image-capturing environment may reappear itself, the evaluated result of fabric smoothness has a good stability.(3) The application software has been programmed. The chief program has a friendly user interface, which mainly covers three functional modularizations, i.e., capturing and reading of fabric image, pretreatment and 3-D reconstruction of the 2-D gray image, extraction and database preservation of the feature parameters, with which fabric image can be captured, analyzed and processed conveniently at real time.(4) The smoothness sorter has been established by using BP nerve network. The sorter takes the American AATCC-124 smoothness replicas as the reference standard, which not only inherits the present international standard for fabric smoothness subjective evaluation, but also realizes digitization of the replicas information, effectively avoiding the artificial influence like the subjective evaluation.(5) The experiment has been carried out on the objective assessment system developed in this thesis by selecting thirty-six actual fabrics. It shows that there are thirty-three samples in all without difference between the subjective evaluation and the objective evaluation with this system, which is 91.6% of the thirty-six actual fabrics, and the correlation coefficient between both methods is 94.60%. The result indicates that this system is valid and usable, especially good to the plain-colored fabric.
Keywords/Search Tags:fabric smoothness, objective assessment system, photometric stereo, feature extraction, BP neural network
PDF Full Text Request
Related items