應用深度學習OCR於兒童閱讀管理;Applying Deep Learning OCR in Children′s Reading Management

NCUIR > College of Electrical Engineering & Computer Science > Executive Master of Communication Engineering > Electronic Thesis & Dissertation > Item 987654321/92924

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/92924

Title:	應用深度學習OCR於兒童閱讀管理;Applying Deep Learning OCR in Children′s Reading Management
Authors:	陳緯庭;Chen, Wei-Ting
Contributors:	通訊工程學系在職專班
Keywords:	光學辨識;深度學習;機器學習;人工智能;Optical Recognition;Deep Learning;Machine Learning;Artificial Intelligence
Date:	2024-01-22
Issue Date:	2024-09-19 16:31:00 (UTC+8)
Publisher:	國立中央大學
Abstract:	本論文探討了在兒童閱讀管理領域中應用深度學習光學字元辨識（OCR）的可能性。具體而言，我們使用了PaddleOCR，透過遷移式學習的方法，建立了一個專門用於辨識兒童讀物封面的模型。這個模型的平均辨識率達到了74.5%。傳統的OCR方法在辨識兒童讀物封面時可能遇到一些挑戰，因為兒童讀物通常具有特殊的字型和圖像風格，這對傳統OCR模型來說可能是一個問題。為了克服這些困難，我們採用了深度學習方法，通過PaddleOCR進行模型訓練，並且使用遷移式學習的技術，使模型能夠更好地適應兒童讀物的特殊特徵。實驗結果顯示，在兒童讀物封面辨識方面，平均辨識率達到74.5%。這一結果表明深度學習方法在兒童閱讀管理中具有潛力，可以為教育和圖書管理領域提供有價值的工具。但本研究仍有進一步的改進空間，包括擴大數據集以提高模型的性能，以及研究其他深度學習技術來進一步優化辨識結果。本論文為應用深度學習OCR於兒童閱讀管理領域提供了有價值的參考，為未來的相關研究提供了基礎。;This paper explores the potential of applying deep learning Optical Character Recognition (OCR) in the field of children′s reading management. Specifically, we utilized PaddleOCR and employed transfer learning to develop a specialized model for recognizing covers of children′s reading materials. The model achieved an average recognition rate of 74.5%. Traditional OCR methods may face challenges when recognizing covers of children′s reading materials due to their unique fonts and image styles. To overcome these difficulties, we adopted a deep learning approach, training the model using PaddleOCR and incorporating transfer learning techniques to adapt the model better to the distinct features of children′s reading materials. Experimental results showed an average recognition rate of 74.5% in the recognition of children′s reading material covers. This outcome indicates the potential of deep learning methods in the field of children′s reading management, offering valuable tools for education and library management. However, there is still room for further improvement in this study, including expanding the dataset to enhance model performance and exploring other deep learning techniques for further optimizing recognition results. This paper provides a valuable reference for the application of deep learning OCR in the domain of children′s reading management, laying the foundation for future related research.
Appears in Collections:	[Executive Master of Communication Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	44	View/Open

社群 sharing

Loading...