Page segmentation of Chinese newspapers
J Xi, J Hu, L Wu - Pattern recognition, 2002 - Elsevier
J Xi, J Hu, L Wu
Pattern recognition, 2002•ElsevierThis paper describes a new bottom-up method for page segmentation of Chinese document
images. Because of some special characteristics of Chinese newspaper documents, many
traditional methods developed for English documents fail in segmenting them correctly.
Based on run-length smoothing algorithm and minimal spanning tree clustering, the
proposed method can resolve the problems of segmenting Chinese documents that differ
from English documents.
images. Because of some special characteristics of Chinese newspaper documents, many
traditional methods developed for English documents fail in segmenting them correctly.
Based on run-length smoothing algorithm and minimal spanning tree clustering, the
proposed method can resolve the problems of segmenting Chinese documents that differ
from English documents.
This paper describes a new bottom-up method for page segmentation of Chinese document images. Because of some special characteristics of Chinese newspaper documents, many traditional methods developed for English documents fail in segmenting them correctly. Based on run-length smoothing algorithm and minimal spanning tree clustering, the proposed method can resolve the problems of segmenting Chinese documents that differ from English documents.
Elsevier
Showing the best result for this search. See all results