| Summary: | 碩士 === 靜宜大學 === 財務工程學系 === 107 === With the advancement of information and the evolution of the times, many tasks have been replaced by various automated data processing processes. Information technology can provide more convenient and fast problem solving methods. For example, image recognition technology can assist face recognition. Applications provide us with a more convenient life.
This study uses image automation processing and digital identification technology to construct an information processing process that can retrieve important digital information such as bank account number on the post office passbook. This technology provides automatic account number configuration. This process can improve work efficiency and reduce manpower load.
This study used Python and openCv package for image preprocessing, important information area capture and digital identification. In the image processing part, Image Smoothing, Equalization, Binarization, Dilate and Erode are used for image preprocessing. The main purpose is to solve the image miscellaneous which may be caused by poor signal source and photo source. Then the contours for ROI (region of interest) are used to retieve area which contained bank account number, and finally we use Tesseract-OCR (Optical Character Recognition) to identify account numbers in the ROI section.
|
|---|
