SingTown AI Vision Technology enables more accurate, stable, and intelligent counting on packaging production lines.

In packaging production lines, traditional infrared or mechanical counting methods are prone to interference from light, speed, and object shapes, resulting in inaccurate counts or response delays.

為提升檢測準確度與自動化水平，包裝設備製造商引入了星瞳的OpenMV智能相機AI圖像識別技術。該系統利用AI機器視覺與圖像識別算法，自動檢測並統計輸送帶上的包裝袋，實現實時信號輸出與協調控制。

The packaging counting system utilises SingTown Technology's OpenMV intelligent camera AI image recognition technology to monitor the quantity of packaging bags in real-time on the operating conveyor belt.

The camera is installed 1.5–2 metres away from the conveyor belt to capture images and perform edge AI recognition of packaging bags measuring approximately 70cm × 40cm. The system utilises computer vision algorithms (edge detection + motion recognition + target tracking) to determine the entry and exit status of the packaging. When a complete package is detected passing through the recognition area, it automatically outputs a counting signal.

The OpenMV intelligent camera facilitates flexible edge deployment, equipped with edge computing capabilities and multiple I/O control interfaces. It can directly transmit pulse signals to PLCs, microcontrollers, or host computers via GPIO or serial ports, achieving production synchronisation and output statistics.

The system response time is below 20 milliseconds, ensuring stable adaptation to varying conveying speeds and lighting conditions.

透過SingTown科技嘅OpenMV智能相機圖像識別方案，包裝生產線由人手或光電計數升級至AI視覺計數，顯著提升檢測準確度、穩定性同生產效率。