关键词: 机器视觉, 误差补偿, 轮廓提取, 同步控制, 智能缝制, 服装

Abstract:

Objective In response to problems such as high labor demand, low automation, and low efficiency in the process of fabric stitching, machine vision, industrial robot, and sewing machine technologies were studied and applied to the design and development of intelligent sewing equipment. By combining these technologies, it helps to improve the absolute positioning accuracy of the industrial robot, as well as the ability to extract fabric contours and ensure the consistency of sewing quality. This has reduced the need for manual labor and has helped drive the transformation and upgrading of the garment sewing industry.

Method Firstly, the absolute positioning kinematics parameter error compensation technology of industrial robot based on wire encoder was studied, so that the absolute positioning accuracy of industrial robot can meet the requirements of sewing task. Secondly, the technology of fabric contour and key points extraction based on Canny operator was studied, which provides data basis for the sewing path plan. Thirdly, the speed synchronization control algorithm of industrial robot and sewing machine based on proportional integral differential (PID) law was developed to solve the problem of irregular stitches caused by the unsynchronization of industrial robot and sewing machine.

Results A double-layer garment piece intelligent sewing equipment consisting of industrial robot, camera, sewing machine, working platform, loading, unloading and other auxiliary mechanisms was developed according to customer needs (Fig. 3,4). The calibration software was developed by applying the optimization compensation technology for the absolute positioning error of the industrial robot based on the wire encoder, which reduced the absolute positioning accuracy error of the central point of the industrial robot tool from the average 1.263 5 mm to 0.128 5 mm. Sewing effect showed the comparison of sewing effects before and after optimization of absolute positioning error. Sewing effect before error compensation shows the sewing stitches prior to error optimization and sewing effect after error compensation shows the sewing stitches after error optimization. It can be seen from sewing effect that the sewing quality of the fabric was greatly improved by adopting the error optimization technology. Using fabric contour extraction technology based on Canny operator, a fabric contour and key point's extraction program based on OpenCV was developed. The original fabric image was a fabric of arbitrary shape, representing various contours of real fabric such as straight line, arc and curve. Fabric contour extraction showed the effect of the contour and key points extraction program, generating different numbers of key points according to different contour curvature. The larger the curvature, the more key points were to be extracted, which was convenient for trajectory planning of the industrial robot and improves the sewing quality. The control block diagram and corresponding control program were constructed by using the speed synchronization control technology of the industrial robot and the sewing machine. The sewing machine needle speed was on open loop control, and its speed was proportional to the Tool Center Point (TCP) speed of the industrial robot. The speed of the cloth feeding wheel of the sewing machine was controlled in a closed loop, which improved the ability of the cloth feeding wheel speed to follow the TCP speed of the industrial robot. Sewing effect and stitches demonstrated the effect before and after the speed synchronization of the industrial robot and the sewing machine, leading to significant improvement of the quality of stitches.

Conclusion Combined with machine vision, absolute positioning error compensation of industrial robot, contour extraction technology, synchronous control technology of manipulator and sewing machine, and this research systematically studied the technical system required for intelligent sewing equipment. The research and development of intelligent sewing equipment has effectively reduced the number of equipment operators, solved the problem of shortage of sewing workers, provided technical accumulation for the automation and intelligence of sewing equipment, and also provided a technical basis for the automation, intelligence and digital transformation of the sewing process in the sewing industry. It is recommended that the sewing equipment should be further improved mainly from the following two aspects, i.e., the 6-axis industrial robot should be changed to Selective Compliance Assembly Robot Arm (SCARA) to reduce equipment costs and improve the absolute positioning accuracy, and a simple and easy-to-use human-computer interface should be designed to facilitate operation and switching sewing tasks.

Key words: machine vision, error compensation, contour extraction, synchronization technology, intelligent sewing