| Summary: | 碩士 === 國立中興大學 === 資訊科學學系 === 83 === Optical character recognition(OCR) has been a subject of great
interest to many people. It is very important in realize the
dream of automatic reading of texts from a document. There are
no perfect OCR for handwritten Chinese characters. Typically, a
OCR system is divided into three stages: preprocessing, feature
extraction, and recognition. Globally speaking, a good OCR
system is only decided by its recognition rate. However, good
result of recognition is dependent on good feature extraction
which is dependent on good preprocessing. In this thesis, we
design and implement every stage. The feature extraction of our
system is based on 4-corner number proposed by Uing-Wu Wang.
|
|---|
