Investigation of the processes required for the automation of stitchmarking in shoe manufacture
This thesis describes a novel approach to the high speed automatic stitchmarking of shoe upper components by integrating an electrographic printer with a shape recognition system. A critical review of recognition system parameters selects the currently known shape parameters which are most suitable for use in a high speed recognition system with the large number of different shoe components found in a typical shoe factory. These are compared with the parameters actually used in the previously developed recognition system to be used for stitchmarking. A discussion of printing technologies suitable for marking shoe materials with computer generated patterns follows. It is concluded that an electrographic printer has the best combination of characteristics. There follows a description of experiments demonstrating xerography on shoe upper materials, and the design of a system integrating a low-cost laser printer to the recognition system which proved the concept of continuous automatic stitchmarking. With this performing satisfactorily, the system was converted to use a high speed printer requiring the use of an advanced graphics processor for handling the data transformations and interfacing with the printer. Modifications to the printer for operation with shoe materials are described, together with the need for a special toner. A full description of the resulting stitchmarking system is given, followed by details of analyses of its performance. Individual chapters are devoted to the accuracy, recognition efficiency, and the timing of the system. A potential bottleneck in determining the orientation of certain difficult shapes is identified, and faster methods for dealing with these are specially investigated. This concludes that the best approach is to optimise the present method and accelerate the calculations by using a more advanced microprocessor. A discussion of the general running of the machine includes details of problems which occurred with the modified printer mechanism and how these were overcome. Finally, suggestions are made for incorporation in an improved system capable of handling larger shapes with a tenfold speed increase.