TY - JOUR
T1 - Offline recognition of chinese handwriting by multifeature and multilevel classification
AU - Tang, Yuan Y.
AU - Tu, Lo Ting
AU - Liu, Jiming
AU - Lee, Seong Whan
AU - Lin, Win Win
AU - Shyu, Ing Shyh
N1 - Funding Information:
This work was supported by research grants received from Research Grant Council (RGC) of Hong Kong and Faculty Research Grant (FRG) of Hong Kong Baptist University. This work was also a partial result of the project no. 35N1300 conducted by the Industrial Technology Research Institute under the sponsorship of the Minister of Economic Affairs, Republic of China, and the Hallym Academy of Science, Hallym University, Korea.
PY - 1998/5
Y1 - 1998/5
N2 - One of the most challenging topics is the recognition of Chinese handwriting, especially offline recognition. In this paper, an offline recognition system based on multifeature and multilevel classification is presented for handwritten Chinese characters. Ten classes of multifeatures, such as peripheral shape features, stroke density features, and stroke direction features, are used in this system. The multilevel classification scheme consists of a group classifier and a five-level character classifier, where two new technologies, overlap clustering and Gaussian distribution selector, are developed. Experiments have been conducted to recognize 5,401 daily-used Chinese characters. The recognition rate is about 90 percent for a unique candidate, and 98 percent for multichoice with 10 candidates.
AB - One of the most challenging topics is the recognition of Chinese handwriting, especially offline recognition. In this paper, an offline recognition system based on multifeature and multilevel classification is presented for handwritten Chinese characters. Ten classes of multifeatures, such as peripheral shape features, stroke density features, and stroke direction features, are used in this system. The multilevel classification scheme consists of a group classifier and a five-level character classifier, where two new technologies, overlap clustering and Gaussian distribution selector, are developed. Experiments have been conducted to recognize 5,401 daily-used Chinese characters. The recognition rate is about 90 percent for a unique candidate, and 98 percent for multichoice with 10 candidates.
KW - Gaussian distribution selector
KW - Multifeature
KW - Multilevel classification
KW - Offline chinese handwriting recognition
KW - Overlap clustering
UR - http://www.scopus.com/inward/record.url?scp=0032072171&partnerID=8YFLogxK
U2 - 10.1109/34.682186
DO - 10.1109/34.682186
M3 - Journal article
AN - SCOPUS:0032072171
SN - 0162-8828
VL - 20
SP - 556
EP - 561
JO - IEEE Transactions on Pattern Analysis and Machine Intelligence
JF - IEEE Transactions on Pattern Analysis and Machine Intelligence
IS - 5
ER -