Abstract
Today, there are many documents in the form of digital images obtained from various sources which must be able to be processed by a computer automatically. One of the document image processing is text feature extraction using OCR (Optical Character Recognition) technology. However, in many cases OCR technology are unable to read text characters in digital images accurately. This could be due to several factor such as poor image quality or noise. In order to get accurate result, the image must be in a good quality, so that digital image need to be preprocessed. The image preprocessing method used in this study are Otsu Thressholding Binarization, Niblack, and Sauvola methods. While the OCR technology used to extract the character is Tesseract library in Python. The test results show that direct text extraction from the original image gives better results with a character match rate average of 77.27%. Meanwhile, the match rate using the Otsu Thressholding method was 70.27%, the Sauvola method was 69.67%, and the Niblack method was only 35.72%. However, in some cases in this research the Sauvola and Otsu methods give better results.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
