隨著深度學習(Deep Learning, DL)之進步,進行物件偵測(Object detection)與分類(Classification)應用也越來越廣。現今製造領域運用深度學習技術越來越多,例如產品識別、預測產能……等。本研究主要針對製造業在進行產品加工時,預防機械作業人員疏失而導致操作安全事件發生。例如,作業人員操作機械設備時,因疏忽將手伸入運作中的機械設備,或是操作機械設備時未依規定配戴防護手套,導致手被機械設備弄傷,故本研究希望能夠透過深度學習技術,於機器設備運作時檢測是否有因規定配戴手套或是在機械設備運作時將手伸入機台內的情形,可應用於防止作業人員操作機械設備不慎導致之工作安全事件發生,提升作業人員操作機械設備工作安全。
With the advancement of deep learning (DL), the application of object detection and classification has become more and more widespread. There are more and more applications of deep learning in today's manufacturing field, such as product identification, production capacity prediction, etc. This research mainly focuses on preventing the occurrence of man-made safety incidents caused by the negligence of operators during product processing in the manufacturing industry. For example, when operators are operating a machinery equipment, they are careless with extending their hands into the machinery and equipment while the machinery is running. During the operation of machinery without wearing protective gloves due to regulations, the hands may injured by the machinery. Therefore, this research studies to detect whether there exists a hand without wearing glove in a running machinery that equips with the proposed deep learning method. Alarming the users who extend their hands into the machine during operation can prevent work safety incidents caused by accidental operation of machinery.