COMPARISON OF IMAGE SEGMENTATION METHOD IN IMAGE CHARACTER EXTRACTION PREPROCESSING USING OPTICAL CHARACTER RECOGINITON

Today, there are many documents in the form of digital images obtained from various sources which must be able to processed by a computer automatically. One document image processing is text feature extraction using OCR (Optical Character Recognition) technology. However, cases technology unable read characters accurately. This could due several factor such as poor quality or noise. In order get accurate result, good quality, so that need preprocessed. The preprocessing method used this study Otsu Thressholding Binarization, Niblack, and Sauvola methods. While extract character Tesseract library Python. test results show direct original gives better with match rate average 77.27%. Meanwhile, was 70.27%, 69.67%, Niblack only 35.72%. some research methods give results.