Experimental control parameters.

This study focuses on developing an automatic high-precision alignment system that integrates automated optical inspection with machine vision recognition to replace traditional manual alignment methods. The core aim is to implement an image-based positioning and control system for a precision alignment platform with dual cameras. Various image processing techniques, such as template matching, edge detection, mathematical morphology, and Hough transforms, are used to identify target features. The identified target coordinates are then fed back to the controller, which calculates the necessary compensation for the XXY platform to achieve precise alignment. For motion control, both XYZ and XXY platforms are utilized to perform image-based alignment tasks. Objects are first transported on the XYZ platform, then precisely positioned and aligned on the XXY platform equipped with dual machine visions, which is equivalent to simulate high-precision automation in production lines. The control system is built on an FPGA development platform integrated with a PC-based architecture and embedded the control algorithms, Proportional-Integral-Derivative (PID) control method and Manifold Deformation Design Scheme (MDDS). Experimental results show that MDDS offers superior performance, compared with PID method, including better accuracy, improved trajectory tracking, and greater control stability. Additionally, the image processing system achieves a high precision, with the flexibility to switch targets, enhancing the system’s practical applicability in real-world scenarios. The proposed system is validated for use in automated production lines, demonstrating its high potential for industrial applications that require precise alignment and control.