花朵無人商店系統

Abstract

中文摘要

隨著勞動力成本上升和技術進步，無人商店在全球迅速發展，Amazon Go和淘咖啡等案例尤為突出。這些無人商店不僅能降低企業成本，還提供了更快速便捷的購物體驗。本研究開發了一套無人花朵商店系統，運用物件偵測技術自動辨識花朵的位置和種類，並結合精心設計的使用者介面，大幅提升顧客的購物體驗。該系統通過蒐集逢甲大學校內花朵的數據進行模型訓練，並使用YOLOv8模型實現高準確度的花朵辨識。實驗結果表明，該系統能夠精確識別四種不同花朵，提供了一種有效的解決方案來改進無人商店的顧客服務和管理。

Abstract With rising labor costs and technological advancements, unmanned stores are rapidly expanding worldwide, exemplified by Amazon Go and Alibaba's Tao Café. These stores not only reduce operational costs for businesses but also provide faster and more convenient shopping experiences. This study developed an unmanned flower store system utilizing object detection technology to automatically identify the location and types of flowers. The system integrates a user-friendly human-machine interface (HMI) designed to significantly enhance customer experience. By collecting flower data from Feng Chia University's campus and training the YOLOv8 model, the system achieved high accuracy in flower recognition. Experimental results demonstrate that the system can accurately identify four different types of flowers, offering an effective solution for improving customer service and management in unmanned stores.