|
KMID : 0380319920500000079
|
|
Journal of Korean Research Institute for Better Living
1992 Volume.50 No. 0 p.79 ~ p.92
|
|
|
Recognition of Hangul Characters Including Noise Margin
|
|
|
|
|
Abstract
|
|
|
Character recognition is one of the fundamental steps in the research area of computer vision. This paper proposes a new technique for the recognition of presegmented Hangul characters. Unlike the early works which are sensitive to noise and shift variation, the proposed approach applying the general two dimensional convolution to learning patterns is rather insensitive to the input images.
In this paper, Hangul recognition based neural network is implemented. Experimental recognition rate is compared in two ways of learning ; with-convolution and without-convolution learning. In our experimental result, Hangul recognition using the prior optimal convolution learning gives a better recall and less sensitive to the input noises.
|
|
|
KEYWORD
|
|
|
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
|
|