SingTown Technology develops an AI-powered visual waste sorting system, offering greater precision, lower costs, and easier deployment.

In the context of smart cities and environmental governance, SingTown Technology's AI smart camera system, based on OpenMV, enables automatic garbage identification and sorting. Utilising machine vision and image recognition algorithms, the system can accurately distinguish various types of waste, such as plastic water bottles, paper balls, batteries, and fruit peels, within a range of approximately 5 metres. It coordinates with robotic arms to achieve automatic sorting and disposal, contributing to the development of smart city and environmental automation systems.

With the deployment of intelligent waste sorting systems across various regions, their limitations have gradually come to light: sorting methods based on infrared or weight recognition offer limited accuracy and still rely on manual intervention, which in turn suffers from low efficiency and high labour intensity. SingTown Technology's intelligent camera solution, built on OpenMV, leverages AI image recognition and computer vision algorithms to deliver a more precise embedded vision solution for intelligent waste identification and automated sorting.

The system employs high-precision cameras to capture real-time images of waste, analysing its shape, colour, and texture features through edge AI to accurately identify typical waste types such as plastic water bottles, crumpled A4 paper, size-7 batteries, and orange peels. The system automatically transmits the results to the robotic arm control module for precise disposal into corresponding waste bins, with the entire process requiring no manual intervention.

此外，OpenMV平台支援用戶自訂模型訓練，能夠根據不同垃圾風格、物料或地區差異，快速優化並適應特定場景。即使在光線不均或背景複雜的環境中，系統仍能保持高識別率及實時反應能力。

此方案嘅邊緣部署優勢，令其能夠廣泛應用於社區垃圾分類中心、校園環保教育、工廠廢物管理同城市智能回收系統。透過整合人工智能、機器視覺同圖像識別技術，SingTown令智能相機成為環保分類嘅核心力量，為建設更高效、潔淨同智能嘅未來城市作出貢獻。