绗缝机花样加工代码自动生成算法

为简化绗缝机花样设计工艺,降低人的经验影响,达到随拍随绣的效果,对目标绗缝图像进行预处理与边缘提取;采用轮廓跟踪算法,将图像分割为若干有序点链轮廓;最后使用遗传算法,以直线和圆弧为微段加工基元,以一阶连续为约束条件,拟合各段轮廓,进行模拟加工,同时运用MatLab进行编程验证,证明其可靠性。结果表明:应用带约束的误差函数,避免了因端点不连续带来的剪线、跳针的影响,各代最优加工路径的适应度在遗传算子的作用下逐步提高,达到总误差小且拟合段数少的效果;模拟加工实时性好、效率高、还原度高,满足实践中复杂花样的绗缝加工需求。

Automatic generation algorithm for pattern processing codes of quilting machines

In order to simplify the pattern design process with the quilting machines, to reduce the influence of operators' experience and to achieve the effect of the real-time quilting, this research worked on the automatic generation of pattern codes. The target quilting pattern is pre-processed and edge-extracted, and the contour tracking algorithm was used to segment the image into several ordered point-chain contours. The genetic algorithm was used with the line and arc as base processing elements and the first-order continuous continuity as constraint condition, to fit each outline contour to carry out simulation machining. Programming verification was carried out to prove the reliability of the result by MatLab. The results show that the constraint error function is able to avoid the influence of the trimming and stitch skipping caused by the discontinuity of the endpoint. The fitness of optimal processing path of each generation is gradually improved by the action of the genetic operators. The total error and the number of fitting segments are smaller. The pattern simulation processing satisfies the requirements of good real-time performance, high efficiency, high degree of reduction and the needs of quilting manufacturing requirements of complex patterns in practice.