Research And Design Of Software For Clothing Cutting Systems

Numerical control cutting of clothing is inevitable trend of clothing processing developmentand the advanced level greatly influences processing and manufacturing of clothing. Theclothing cutting equipment is developing to the quicker and more accurate direction and controlsoftware of clothing cutting machine is the soul of system, which has profound significance toimprove performance of cutting machine, labor capacity of clothing industry and competitivepower of enterprise.It designs platform of software control about intelligent clothing cutting system on the basisof analyzing clothing cutting technology in and abroad as well as present and trend ofdevelopment about cutting equipment. It has the embedded platform based on Linux and usesindustrial camera to input original image information, which can realize intelligent handing ofprocessing image, analysis of PLT file and real time display of processing trajectory, at the sametime, it realizes intelligence, internetization and systematization to the whole software controlsystem through network connection to realize processing data share and remote network monitorand fault detection. The main studies and conclusions of this paper are listed as below:Firstly, it selects embedded hardware platform and transplants U-BOOT, Linux kernel, filesystem and Qt dynamic library according to requirement of system function. At the same time, itbuilds platform of software development and environment of cross compiler to convenientdevelopment and debugging of software.Secondly, it adopts model of camera in OpenCV to standardize industrial camera and doesimage processing to the corrected image in order to extract contour curve. We should reducesharp point to the best of our ability in contour tracking process to ensure to reduce subsectionnumber in the further handing process to boundary curve.Thirdly, it realizes contour curve vectorization through feature point extraction and curvefitting. It designs a intermediate value and determines corresponding feature point qualitativelyby comparing the intermediate value and threshold to minimize calculation; It fits contour curveof sample about each subsection through quadratic parabolas recursion and re-disperse the fitting curve to produce PLT file.Fourthly, it fits processing data through cubic quasi-uniform B-sample and disperses fittingcurve by adopting arc-line approximation, which can improve PLT file not produced by imageand reduce capacity of stored data.Fifthly, it designs PLT files in downloading network communication and controls DSPprocessing through serial communication.Finally, it is summarized the research work and proposed to be improved.