每50毫秒精準校正，SingTown科技令布料檢測更智能。

In the production process of textile fabrics, the warp and weft threads, bright filaments, or reflective textures on the fabric surface must maintain strict straightness and positional consistency.
當高速輸送布料出現亮線偏差或多線重疊時，容易影響後續印刷、壓花或塗層工序的對位準確度。
為此，紡織製造企業採用SingTown Technology的OpenMV智能相機識別技術，結合AI機器視覺與動態線條檢測算法，實現對高速移動布料上亮線位置的實時檢測與偏移量測量。
This system is applicable to production processes such as textile weaving, printing and dyeing, and fabric quality inspection, enabling online automatic detection and intelligent deviation correction control.

紡織製造企業基於SingTown Technology嘅OpenMV智能相機識別技術，實現咗亮線偏差嘅高精度檢測同自動閉環修正系統。

此方案涉及在生产线上固定安装高速工业相机，以捕捉布料移动方向的垂直影像。当布料以每分钟80至100米的速度通过检测区域时，相机会实时采集布料表面的图像。

The system's computer vision and image recognition technology can extract the bright line features on the fabric surface by utilising brightness detection, edge enhancement, and Hough Transform for line detection.

The AI algorithm determines the centre position of the bright line via brightness gradient analysis and pixel projection techniques, compares it with the pre-set reference position, and provides offset data in real time.

If the bright line exhibits lateral deviation or breakage, the system immediately transmits the coordinate deviation value to the control system, which drives the rapid adjustment of the correction device.

Under fixed lighting and camera conditions, the OpenMV module can suppress motion blur through high-speed exposure, ensuring image clarity and detection stability.

The system's edge AI can independently complete recognition tasks locally, with the entire recognition cycle taking less than 50 milliseconds, meeting the real-time detection requirements for continuously moving fabrics.

透過此解決方案，企業實現了對亮線偏差的高精度檢測與自動修正控制，顯著提升了生產一致性與布料品質。SingTown Technology的OpenMV為智能紡織製造提供了可靠的嵌入式視覺檢測解決方案。