Font Size: a A A

Research Of Algorithm And Software Of Information Extraction From Pattern Drawing Of Jacquard Knitting Machine

Posted on:2008-11-09Degree:MasterType:Thesis
Country:ChinaCandidate:J LiFull Text:PDF
GTID:2178360245997804Subject:Instrument Science and Technology
Abstract/Summary:PDF Full Text Request
With the development of science technology, computer plays a more and more important role in textile industry. Though the jacquard knitting machine controlled by computer improves the automation level of textile industry, there still exist problems in the manual input of textile pattern drawing, such as heavy workload, low efficiency and high mistake rate, etc. Addressing these problems, we proposed an automatic extraction algorithm of pattern information and developed the software Lace Pattern Designer.According to the results of analyzing image feature, we choose the threshold of image based upon Da-Jing method, then detected line and point object of image, finally obtained the relative coordinates of the point object from the grid. The detection of grid lines combines optimized LOG edge detectors and two modified Hough transformations. The identification of point objects depends on mathematical morphology and connected-components labeling. Experimental results showed that the accuracy of the algorithm was superior to 96%.In the paper, we analyzed the feasibility of the software, designed software development model and compared several Object Oriented development environment, then developed the software in Visual C++ environment. The software applies the Single Document Interface supplied by the Microsoft Foundation Class and the core of the procedure is the automatic extraction algorithm. Designers can use the software to extract the information from drawing, design pattern in grid, compare the results, print drawings and so on. Finally, the late work includes software encryption, making help and installation document. The test of software indicated its functions achieved the prearranged development goals, namely greatly saving the time that designer used to design or input a pattern drawing, and significantly reducing probability of mistakes.This software can automatically extract pattern information from grid images, solve the problem in the manual input of textile pattern drawing. The software has reasonable design, versatile function, direct views, clear operational procedure and easy to learn. The application of the software justified its validity.
Keywords/Search Tags:Feature extraction, Line detection, Hough transformations, connected-components labeling, textile pattern drawing
PDF Full Text Request
Related items