頁籤選單縮合
題 名 | 基於卷積神經網路之校園緊急求助系統=Campus Emergency Call System Based on Convolutional Neural Network |
---|---|
作 者 | 謝依蓓; 李國彰; 陳思明; 黃皇鳴; 黃承恩; 吳家瑋; 林宛諭; | 書刊名 | 慈濟科技大學學報 |
卷 期 | 13=37 2024.03[民113.03] |
頁 次 | 頁125-142 |
分類號 | 312.83 |
關鍵詞 | 人工智慧; 影像辨識; 卷積神經網路; 校園安全; Artificial intelligence; Image recognition; Convolutional neural network; Campus security; |
語 文 | 中文(Chinese) |
中文摘要 | 近年來基於卷積神經網路技術被廣泛用於人工智慧模型訓練,例如:偵測與辨識系統、監控系統等。本論文透過人工智慧模型結合電腦視覺,進行環境影像偵測,透過多層卷積網路神經元架構學習校園地點辨識。希望藉由人工智慧結合緊急通報求助,快速且精確找出發生意外的地點,通知校園警衛,解決突發事件。在實驗中,我們以手機拍攝不同教室照片做測試,傳輸至人工智慧系統,進行辨識與定位事故發生地點,辨識結果再發送至警衛端,緊急通知警衛至事發地點進行查看。求助者能於校園內任何地方進行求助,藉由人工智慧影像辨識定位,可以解決一些問題,例如:GPS受周遭環境干擾,無法確認地點問題。與Wi-Fi設備普及度不夠,而無法精準定位地點問題。 |
英文摘要 | Recent years, convolutional neural network-based technology has been widely used in artificial intelligence model training and applied to various artificial intelligence systems, such as: detection and identification systems, monitoring systems, authentication systems, etc. This paper uses artificial intelligence model training combined with computer vision to do environmental image detection. We use our own collection of campus photos to train a multi-layer convolutional neural network model with location identification function. In the system, we can quickly and accurately find out the location of the accident, notify the campus security guard, solve emergencies and reduce the harm suffered by help-seekers. In the experiment, we used our mobile phone to take photos of different classrooms, and transmitted photos to the artificial intelligence system to identify and locate the location of the accident. The identification results were sent to the security guard immediately. Help seekers can seek help at any place on campus. |
本系統中英文摘要資訊取自各篇刊載內容。