Research On Eliminating Set Marks By Image Processing Technique

Set marks is an important factor that leading to woven defects. It is impossible to mend set marks once they are formed. Therefore, set marks can bring to a decrease of fabric quality and give a negative influence on weaving factories. Furthermore, set marks are even more serious when using modern high-speed looms, especially when thin fabrics are woven. Researchers have been searching for various methods to eliminate set marks for a long time. However, though many solutions have been proposed, the problem of set marks has not been resolved completely.A new scheme to eliminate set marks based on detecting cloth fell displacement by untouched image processing is proposed in this paper. Accurate detection of cloth fell displacement is a key to this scheme. New image processing ways that can greatly improve the calculating speed is designed, they are SSDA algorithem based Mallat decomposition and template matching algorithem based Mallat decomposition. A set of experiment equipments are designed to detect the cloth-fell displacement. The image of cloth which closes to the cloth-fell can be grabbed by a CCD camera and processed by a computer on line, the cloth fell displacement is precisely calculated by self-developed computer program. Because fabric that under tension will creep and slack during some time, experiments are performed to probe this change using the self-designed equipments, and the results are analyzed. After that, experiments are performed on a real loom, and the experiments results are analyzed. The experiment results prove that, the approach we developed can detect the cloth-fell displacement accurately. Therefore, integrating the approach into modern weaving loom, set markscould be eliminated completely.Chapter 1 introduces the development and application of digital image processing, especially in textile industry.Chapter 2 describes the related image processing technologies.In Chapter 3, a new scheme that eliminates cloth-fell displacement based on image processing technology is presented, involving the required hardware and software environment in the lab.Chapter 4 presents theories and applications of image matching technology, and some algorithms of image matching in detail.The application and relevant theories of wavelet transform are described in Chapter 5, it mainly focus on the construction of biorthogonal wavelet base and mallat algorithm application to digital image decompose.In Chapter 6, two new algorithms are advanced and designed, and the programs developed in VC which integrates the proposed algorithms are used to detect the cloth-fell displacement on the lab equipments and loom. The results are satisfying.