http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
이형렬 김천과학대학 1994 김천과학대학 논문집 Vol.20 No.-
This paper presented a method to recognize koran characters using the relational information of stroke location. In the proposed method, feature points are extracted by carrying out preprocessing step such as thinning and the features of strokes are extracted by four directional code. And then, the feature vector that contains the relational information of stroke location is composed and it is compared with feature vectors in the vector table to recognize character. Experimental results show that the recognition rates printed Hangeul are 92% respectively.
이형렬 김천과학대학 1993 김천과학대학 논문집 Vol.19 No.-
This paper presented a method to recognize characters using the location of features. In this method, features are extracted from thinned image and the feature vector that contains the information of feature location is composed and used to recognize character. Experimental results show that the recognition rates for handwritten digit and printed Hangeul are 92% and 93%,respectively.
이형렬 김천과학대학 1995 김천과학대학 논문집 Vol.21 No.-
A thinning algorithm is a vary important factor in order to increase recognition rate in the character recognition. This paper is the study of a parpllel thinning algorithm available for the recognition of various characters, and it proposes the thinning algorithm which can extract the skeleton of the recognized character almost similar to the original pattern by using the depth value representing distance from boundry. Characteristics of the proposed algorithm are easiness of implementation of parallelism, the result of thinning is one pixel width. The proposed algorithm requires only a single pass per iteration; hence, it is substantially faster than other algorithms.